diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100755--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25256 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3508515815085158,
+  "eval_steps": 500,
+  "global_step": 3605,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 9.732360097323601e-05,
+      "grad_norm": 16.13226202148005,
+      "learning_rate": 3.2362459546925574e-08,
+      "loss": 1.1997,
+      "step": 1
+    },
+    {
+      "epoch": 0.00019464720194647202,
+      "grad_norm": 15.765097511926365,
+      "learning_rate": 6.472491909385115e-08,
+      "loss": 1.384,
+      "step": 2
+    },
+    {
+      "epoch": 0.00029197080291970805,
+      "grad_norm": 16.64113665586635,
+      "learning_rate": 9.70873786407767e-08,
+      "loss": 1.2291,
+      "step": 3
+    },
+    {
+      "epoch": 0.00038929440389294404,
+      "grad_norm": 20.34864047521242,
+      "learning_rate": 1.294498381877023e-07,
+      "loss": 0.9025,
+      "step": 4
+    },
+    {
+      "epoch": 0.00048661800486618007,
+      "grad_norm": 28.710711096046108,
+      "learning_rate": 1.6181229773462782e-07,
+      "loss": 1.0305,
+      "step": 5
+    },
+    {
+      "epoch": 0.0005839416058394161,
+      "grad_norm": 21.945801992582915,
+      "learning_rate": 1.941747572815534e-07,
+      "loss": 1.0979,
+      "step": 6
+    },
+    {
+      "epoch": 0.0006812652068126521,
+      "grad_norm": 23.947905905644966,
+      "learning_rate": 2.26537216828479e-07,
+      "loss": 1.1909,
+      "step": 7
+    },
+    {
+      "epoch": 0.0007785888077858881,
+      "grad_norm": 19.835016686730835,
+      "learning_rate": 2.588996763754046e-07,
+      "loss": 1.2083,
+      "step": 8
+    },
+    {
+      "epoch": 0.0008759124087591241,
+      "grad_norm": 16.926846352507788,
+      "learning_rate": 2.9126213592233014e-07,
+      "loss": 1.2369,
+      "step": 9
+    },
+    {
+      "epoch": 0.0009732360097323601,
+      "grad_norm": 21.349924470647284,
+      "learning_rate": 3.2362459546925565e-07,
+      "loss": 1.0052,
+      "step": 10
+    },
+    {
+      "epoch": 0.0010705596107055961,
+      "grad_norm": 25.127579741628022,
+      "learning_rate": 3.5598705501618125e-07,
+      "loss": 1.2631,
+      "step": 11
+    },
+    {
+      "epoch": 0.0011678832116788322,
+      "grad_norm": 12.524049131196549,
+      "learning_rate": 3.883495145631068e-07,
+      "loss": 1.0884,
+      "step": 12
+    },
+    {
+      "epoch": 0.001265206812652068,
+      "grad_norm": 20.706648432487587,
+      "learning_rate": 4.207119741100324e-07,
+      "loss": 1.1469,
+      "step": 13
+    },
+    {
+      "epoch": 0.0013625304136253042,
+      "grad_norm": 17.655230197318655,
+      "learning_rate": 4.53074433656958e-07,
+      "loss": 1.2922,
+      "step": 14
+    },
+    {
+      "epoch": 0.00145985401459854,
+      "grad_norm": 16.550170455008725,
+      "learning_rate": 4.854368932038835e-07,
+      "loss": 1.1792,
+      "step": 15
+    },
+    {
+      "epoch": 0.0015571776155717761,
+      "grad_norm": 24.456798845425887,
+      "learning_rate": 5.177993527508092e-07,
+      "loss": 1.0804,
+      "step": 16
+    },
+    {
+      "epoch": 0.0016545012165450122,
+      "grad_norm": 14.659117460865279,
+      "learning_rate": 5.501618122977346e-07,
+      "loss": 1.0973,
+      "step": 17
+    },
+    {
+      "epoch": 0.0017518248175182481,
+      "grad_norm": 15.324823146378344,
+      "learning_rate": 5.825242718446603e-07,
+      "loss": 0.9791,
+      "step": 18
+    },
+    {
+      "epoch": 0.0018491484184914842,
+      "grad_norm": 12.483869597287145,
+      "learning_rate": 6.148867313915858e-07,
+      "loss": 1.0829,
+      "step": 19
+    },
+    {
+      "epoch": 0.0019464720194647203,
+      "grad_norm": 11.921211994178957,
+      "learning_rate": 6.472491909385113e-07,
+      "loss": 0.6862,
+      "step": 20
+    },
+    {
+      "epoch": 0.0020437956204379564,
+      "grad_norm": 14.53279456676939,
+      "learning_rate": 6.79611650485437e-07,
+      "loss": 0.7814,
+      "step": 21
+    },
+    {
+      "epoch": 0.0021411192214111923,
+      "grad_norm": 15.68359520937104,
+      "learning_rate": 7.119741100323625e-07,
+      "loss": 0.883,
+      "step": 22
+    },
+    {
+      "epoch": 0.002238442822384428,
+      "grad_norm": 14.062468532950906,
+      "learning_rate": 7.443365695792882e-07,
+      "loss": 1.0087,
+      "step": 23
+    },
+    {
+      "epoch": 0.0023357664233576644,
+      "grad_norm": 11.150778403716444,
+      "learning_rate": 7.766990291262136e-07,
+      "loss": 0.4884,
+      "step": 24
+    },
+    {
+      "epoch": 0.0024330900243309003,
+      "grad_norm": 7.740982223602688,
+      "learning_rate": 8.090614886731392e-07,
+      "loss": 0.8543,
+      "step": 25
+    },
+    {
+      "epoch": 0.002530413625304136,
+      "grad_norm": 6.4338060169141915,
+      "learning_rate": 8.414239482200648e-07,
+      "loss": 0.7948,
+      "step": 26
+    },
+    {
+      "epoch": 0.002627737226277372,
+      "grad_norm": 6.227022582398367,
+      "learning_rate": 8.737864077669904e-07,
+      "loss": 0.7814,
+      "step": 27
+    },
+    {
+      "epoch": 0.0027250608272506084,
+      "grad_norm": 7.989531820662516,
+      "learning_rate": 9.06148867313916e-07,
+      "loss": 0.5645,
+      "step": 28
+    },
+    {
+      "epoch": 0.0028223844282238442,
+      "grad_norm": 6.4745089193753,
+      "learning_rate": 9.385113268608415e-07,
+      "loss": 0.6802,
+      "step": 29
+    },
+    {
+      "epoch": 0.00291970802919708,
+      "grad_norm": 8.23650018531745,
+      "learning_rate": 9.70873786407767e-07,
+      "loss": 0.6218,
+      "step": 30
+    },
+    {
+      "epoch": 0.0030170316301703164,
+      "grad_norm": 4.915479010119541,
+      "learning_rate": 1.0032362459546926e-06,
+      "loss": 0.8879,
+      "step": 31
+    },
+    {
+      "epoch": 0.0031143552311435523,
+      "grad_norm": 4.288138757396447,
+      "learning_rate": 1.0355987055016184e-06,
+      "loss": 0.5917,
+      "step": 32
+    },
+    {
+      "epoch": 0.003211678832116788,
+      "grad_norm": 4.230901102531741,
+      "learning_rate": 1.0679611650485437e-06,
+      "loss": 0.7373,
+      "step": 33
+    },
+    {
+      "epoch": 0.0033090024330900245,
+      "grad_norm": 4.714303656539792,
+      "learning_rate": 1.1003236245954693e-06,
+      "loss": 0.5886,
+      "step": 34
+    },
+    {
+      "epoch": 0.0034063260340632603,
+      "grad_norm": 4.1204943469600925,
+      "learning_rate": 1.132686084142395e-06,
+      "loss": 0.5991,
+      "step": 35
+    },
+    {
+      "epoch": 0.0035036496350364962,
+      "grad_norm": 3.124375547961107,
+      "learning_rate": 1.1650485436893206e-06,
+      "loss": 0.432,
+      "step": 36
+    },
+    {
+      "epoch": 0.0036009732360097325,
+      "grad_norm": 3.741153837090354,
+      "learning_rate": 1.197411003236246e-06,
+      "loss": 0.6379,
+      "step": 37
+    },
+    {
+      "epoch": 0.0036982968369829684,
+      "grad_norm": 3.7740270813504506,
+      "learning_rate": 1.2297734627831717e-06,
+      "loss": 0.5595,
+      "step": 38
+    },
+    {
+      "epoch": 0.0037956204379562043,
+      "grad_norm": 4.783986424289694,
+      "learning_rate": 1.2621359223300972e-06,
+      "loss": 0.8717,
+      "step": 39
+    },
+    {
+      "epoch": 0.0038929440389294406,
+      "grad_norm": 4.242597978097827,
+      "learning_rate": 1.2944983818770226e-06,
+      "loss": 0.6632,
+      "step": 40
+    },
+    {
+      "epoch": 0.0039902676399026765,
+      "grad_norm": 4.309602952976607,
+      "learning_rate": 1.3268608414239483e-06,
+      "loss": 0.7191,
+      "step": 41
+    },
+    {
+      "epoch": 0.004087591240875913,
+      "grad_norm": 4.136462382872819,
+      "learning_rate": 1.359223300970874e-06,
+      "loss": 0.6782,
+      "step": 42
+    },
+    {
+      "epoch": 0.004184914841849148,
+      "grad_norm": 4.2148643401229,
+      "learning_rate": 1.3915857605177997e-06,
+      "loss": 0.8932,
+      "step": 43
+    },
+    {
+      "epoch": 0.0042822384428223845,
+      "grad_norm": 3.829331188520966,
+      "learning_rate": 1.423948220064725e-06,
+      "loss": 0.4697,
+      "step": 44
+    },
+    {
+      "epoch": 0.004379562043795621,
+      "grad_norm": 3.4564347781684557,
+      "learning_rate": 1.4563106796116506e-06,
+      "loss": 0.3377,
+      "step": 45
+    },
+    {
+      "epoch": 0.004476885644768856,
+      "grad_norm": 3.319649807488789,
+      "learning_rate": 1.4886731391585763e-06,
+      "loss": 0.4589,
+      "step": 46
+    },
+    {
+      "epoch": 0.0045742092457420926,
+      "grad_norm": 3.8856546910308034,
+      "learning_rate": 1.5210355987055017e-06,
+      "loss": 0.8413,
+      "step": 47
+    },
+    {
+      "epoch": 0.004671532846715329,
+      "grad_norm": 3.7955924171570605,
+      "learning_rate": 1.5533980582524272e-06,
+      "loss": 0.588,
+      "step": 48
+    },
+    {
+      "epoch": 0.004768856447688564,
+      "grad_norm": 4.5762685715882805,
+      "learning_rate": 1.585760517799353e-06,
+      "loss": 0.6472,
+      "step": 49
+    },
+    {
+      "epoch": 0.004866180048661801,
+      "grad_norm": 4.284420204063246,
+      "learning_rate": 1.6181229773462783e-06,
+      "loss": 0.5233,
+      "step": 50
+    },
+    {
+      "epoch": 0.004963503649635037,
+      "grad_norm": 4.0399534913964645,
+      "learning_rate": 1.650485436893204e-06,
+      "loss": 0.6737,
+      "step": 51
+    },
+    {
+      "epoch": 0.005060827250608272,
+      "grad_norm": 4.850258079033273,
+      "learning_rate": 1.6828478964401297e-06,
+      "loss": 0.5017,
+      "step": 52
+    },
+    {
+      "epoch": 0.005158150851581509,
+      "grad_norm": 3.289730774319516,
+      "learning_rate": 1.715210355987055e-06,
+      "loss": 0.6378,
+      "step": 53
+    },
+    {
+      "epoch": 0.005255474452554744,
+      "grad_norm": 3.116783938182044,
+      "learning_rate": 1.7475728155339808e-06,
+      "loss": 0.5681,
+      "step": 54
+    },
+    {
+      "epoch": 0.00535279805352798,
+      "grad_norm": 3.5896487509946677,
+      "learning_rate": 1.7799352750809063e-06,
+      "loss": 0.5222,
+      "step": 55
+    },
+    {
+      "epoch": 0.005450121654501217,
+      "grad_norm": 3.3627737905222146,
+      "learning_rate": 1.812297734627832e-06,
+      "loss": 0.351,
+      "step": 56
+    },
+    {
+      "epoch": 0.005547445255474452,
+      "grad_norm": 3.405981770724818,
+      "learning_rate": 1.8446601941747574e-06,
+      "loss": 0.5832,
+      "step": 57
+    },
+    {
+      "epoch": 0.0056447688564476885,
+      "grad_norm": 3.231134680455488,
+      "learning_rate": 1.877022653721683e-06,
+      "loss": 0.5558,
+      "step": 58
+    },
+    {
+      "epoch": 0.005742092457420925,
+      "grad_norm": 4.2963387449464605,
+      "learning_rate": 1.9093851132686085e-06,
+      "loss": 0.7544,
+      "step": 59
+    },
+    {
+      "epoch": 0.00583941605839416,
+      "grad_norm": 3.3678084152804315,
+      "learning_rate": 1.941747572815534e-06,
+      "loss": 0.554,
+      "step": 60
+    },
+    {
+      "epoch": 0.0059367396593673965,
+      "grad_norm": 3.635756089652443,
+      "learning_rate": 1.9741100323624596e-06,
+      "loss": 0.5312,
+      "step": 61
+    },
+    {
+      "epoch": 0.006034063260340633,
+      "grad_norm": 3.91764256649437,
+      "learning_rate": 2.006472491909385e-06,
+      "loss": 0.4329,
+      "step": 62
+    },
+    {
+      "epoch": 0.006131386861313868,
+      "grad_norm": 3.4866607421863565,
+      "learning_rate": 2.0388349514563107e-06,
+      "loss": 0.4453,
+      "step": 63
+    },
+    {
+      "epoch": 0.006228710462287105,
+      "grad_norm": 2.9369425143161147,
+      "learning_rate": 2.0711974110032367e-06,
+      "loss": 0.467,
+      "step": 64
+    },
+    {
+      "epoch": 0.006326034063260341,
+      "grad_norm": 3.0906723589687024,
+      "learning_rate": 2.103559870550162e-06,
+      "loss": 0.3917,
+      "step": 65
+    },
+    {
+      "epoch": 0.006423357664233576,
+      "grad_norm": 3.5121616512799747,
+      "learning_rate": 2.1359223300970874e-06,
+      "loss": 0.6428,
+      "step": 66
+    },
+    {
+      "epoch": 0.006520681265206813,
+      "grad_norm": 3.470270871630247,
+      "learning_rate": 2.1682847896440134e-06,
+      "loss": 0.586,
+      "step": 67
+    },
+    {
+      "epoch": 0.006618004866180049,
+      "grad_norm": 2.8689679430782498,
+      "learning_rate": 2.2006472491909385e-06,
+      "loss": 0.2938,
+      "step": 68
+    },
+    {
+      "epoch": 0.006715328467153284,
+      "grad_norm": 4.115573400175418,
+      "learning_rate": 2.2330097087378645e-06,
+      "loss": 0.3855,
+      "step": 69
+    },
+    {
+      "epoch": 0.006812652068126521,
+      "grad_norm": 3.903319335204406,
+      "learning_rate": 2.26537216828479e-06,
+      "loss": 0.6272,
+      "step": 70
+    },
+    {
+      "epoch": 0.006909975669099757,
+      "grad_norm": 2.649165320750572,
+      "learning_rate": 2.297734627831715e-06,
+      "loss": 0.5229,
+      "step": 71
+    },
+    {
+      "epoch": 0.0070072992700729924,
+      "grad_norm": 2.8543884488184235,
+      "learning_rate": 2.330097087378641e-06,
+      "loss": 0.4006,
+      "step": 72
+    },
+    {
+      "epoch": 0.007104622871046229,
+      "grad_norm": 2.9817247056794134,
+      "learning_rate": 2.3624595469255667e-06,
+      "loss": 0.2331,
+      "step": 73
+    },
+    {
+      "epoch": 0.007201946472019465,
+      "grad_norm": 3.592880940053797,
+      "learning_rate": 2.394822006472492e-06,
+      "loss": 0.4889,
+      "step": 74
+    },
+    {
+      "epoch": 0.0072992700729927005,
+      "grad_norm": 2.89844013224274,
+      "learning_rate": 2.427184466019418e-06,
+      "loss": 0.4711,
+      "step": 75
+    },
+    {
+      "epoch": 0.007396593673965937,
+      "grad_norm": 2.6071345596032134,
+      "learning_rate": 2.4595469255663434e-06,
+      "loss": 0.4844,
+      "step": 76
+    },
+    {
+      "epoch": 0.007493917274939173,
+      "grad_norm": 2.9053930844585776,
+      "learning_rate": 2.491909385113269e-06,
+      "loss": 0.5163,
+      "step": 77
+    },
+    {
+      "epoch": 0.0075912408759124085,
+      "grad_norm": 3.4016540038418115,
+      "learning_rate": 2.5242718446601945e-06,
+      "loss": 0.5852,
+      "step": 78
+    },
+    {
+      "epoch": 0.007688564476885645,
+      "grad_norm": 2.7133170026932887,
+      "learning_rate": 2.55663430420712e-06,
+      "loss": 0.4934,
+      "step": 79
+    },
+    {
+      "epoch": 0.007785888077858881,
+      "grad_norm": 3.2321439410345585,
+      "learning_rate": 2.588996763754045e-06,
+      "loss": 0.62,
+      "step": 80
+    },
+    {
+      "epoch": 0.007883211678832117,
+      "grad_norm": 2.6835948161160545,
+      "learning_rate": 2.621359223300971e-06,
+      "loss": 0.4689,
+      "step": 81
+    },
+    {
+      "epoch": 0.007980535279805353,
+      "grad_norm": 4.716894934604404,
+      "learning_rate": 2.6537216828478967e-06,
+      "loss": 0.3364,
+      "step": 82
+    },
+    {
+      "epoch": 0.00807785888077859,
+      "grad_norm": 2.6507857723180646,
+      "learning_rate": 2.686084142394822e-06,
+      "loss": 0.3785,
+      "step": 83
+    },
+    {
+      "epoch": 0.008175182481751826,
+      "grad_norm": 2.356714630861861,
+      "learning_rate": 2.718446601941748e-06,
+      "loss": 0.2591,
+      "step": 84
+    },
+    {
+      "epoch": 0.00827250608272506,
+      "grad_norm": 2.755477478688418,
+      "learning_rate": 2.7508090614886734e-06,
+      "loss": 0.4762,
+      "step": 85
+    },
+    {
+      "epoch": 0.008369829683698296,
+      "grad_norm": 3.7771581783688837,
+      "learning_rate": 2.7831715210355993e-06,
+      "loss": 0.4627,
+      "step": 86
+    },
+    {
+      "epoch": 0.008467153284671533,
+      "grad_norm": 2.8568450908810257,
+      "learning_rate": 2.8155339805825245e-06,
+      "loss": 0.4322,
+      "step": 87
+    },
+    {
+      "epoch": 0.008564476885644769,
+      "grad_norm": 2.914756058289183,
+      "learning_rate": 2.84789644012945e-06,
+      "loss": 0.4835,
+      "step": 88
+    },
+    {
+      "epoch": 0.008661800486618005,
+      "grad_norm": 2.414182197047686,
+      "learning_rate": 2.880258899676376e-06,
+      "loss": 0.493,
+      "step": 89
+    },
+    {
+      "epoch": 0.008759124087591242,
+      "grad_norm": 2.8597853736106975,
+      "learning_rate": 2.912621359223301e-06,
+      "loss": 0.6063,
+      "step": 90
+    },
+    {
+      "epoch": 0.008856447688564476,
+      "grad_norm": 2.4567808863650007,
+      "learning_rate": 2.9449838187702267e-06,
+      "loss": 0.5874,
+      "step": 91
+    },
+    {
+      "epoch": 0.008953771289537713,
+      "grad_norm": 2.819434031784131,
+      "learning_rate": 2.9773462783171527e-06,
+      "loss": 0.552,
+      "step": 92
+    },
+    {
+      "epoch": 0.009051094890510949,
+      "grad_norm": 1.9840396387462764,
+      "learning_rate": 3.0097087378640778e-06,
+      "loss": 0.3736,
+      "step": 93
+    },
+    {
+      "epoch": 0.009148418491484185,
+      "grad_norm": 2.52047300259283,
+      "learning_rate": 3.0420711974110033e-06,
+      "loss": 0.407,
+      "step": 94
+    },
+    {
+      "epoch": 0.009245742092457421,
+      "grad_norm": 3.140839526692518,
+      "learning_rate": 3.0744336569579293e-06,
+      "loss": 0.6513,
+      "step": 95
+    },
+    {
+      "epoch": 0.009343065693430658,
+      "grad_norm": 3.1368865731879554,
+      "learning_rate": 3.1067961165048544e-06,
+      "loss": 0.4804,
+      "step": 96
+    },
+    {
+      "epoch": 0.009440389294403892,
+      "grad_norm": 2.6987222968513196,
+      "learning_rate": 3.13915857605178e-06,
+      "loss": 0.4228,
+      "step": 97
+    },
+    {
+      "epoch": 0.009537712895377129,
+      "grad_norm": 2.5779408707034026,
+      "learning_rate": 3.171521035598706e-06,
+      "loss": 0.4654,
+      "step": 98
+    },
+    {
+      "epoch": 0.009635036496350365,
+      "grad_norm": 2.5189587792888934,
+      "learning_rate": 3.2038834951456315e-06,
+      "loss": 0.5465,
+      "step": 99
+    },
+    {
+      "epoch": 0.009732360097323601,
+      "grad_norm": 2.457408493992738,
+      "learning_rate": 3.2362459546925567e-06,
+      "loss": 0.5077,
+      "step": 100
+    },
+    {
+      "epoch": 0.009829683698296838,
+      "grad_norm": 2.445932328031196,
+      "learning_rate": 3.2686084142394826e-06,
+      "loss": 0.492,
+      "step": 101
+    },
+    {
+      "epoch": 0.009927007299270074,
+      "grad_norm": 2.3199141960061915,
+      "learning_rate": 3.300970873786408e-06,
+      "loss": 0.4432,
+      "step": 102
+    },
+    {
+      "epoch": 0.010024330900243308,
+      "grad_norm": 3.88769555780582,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.3684,
+      "step": 103
+    },
+    {
+      "epoch": 0.010121654501216545,
+      "grad_norm": 2.63905676146042,
+      "learning_rate": 3.3656957928802593e-06,
+      "loss": 0.4238,
+      "step": 104
+    },
+    {
+      "epoch": 0.010218978102189781,
+      "grad_norm": 3.0073749174392885,
+      "learning_rate": 3.398058252427185e-06,
+      "loss": 0.4655,
+      "step": 105
+    },
+    {
+      "epoch": 0.010316301703163017,
+      "grad_norm": 2.613524831872459,
+      "learning_rate": 3.43042071197411e-06,
+      "loss": 0.4948,
+      "step": 106
+    },
+    {
+      "epoch": 0.010413625304136254,
+      "grad_norm": 2.4293628733346764,
+      "learning_rate": 3.462783171521036e-06,
+      "loss": 0.3717,
+      "step": 107
+    },
+    {
+      "epoch": 0.010510948905109488,
+      "grad_norm": 3.3036504610837016,
+      "learning_rate": 3.4951456310679615e-06,
+      "loss": 0.4939,
+      "step": 108
+    },
+    {
+      "epoch": 0.010608272506082725,
+      "grad_norm": 2.6808221933664846,
+      "learning_rate": 3.5275080906148866e-06,
+      "loss": 0.4809,
+      "step": 109
+    },
+    {
+      "epoch": 0.01070559610705596,
+      "grad_norm": 2.853958419293739,
+      "learning_rate": 3.5598705501618126e-06,
+      "loss": 0.4066,
+      "step": 110
+    },
+    {
+      "epoch": 0.010802919708029197,
+      "grad_norm": 5.3412930378250145,
+      "learning_rate": 3.592233009708738e-06,
+      "loss": 0.3599,
+      "step": 111
+    },
+    {
+      "epoch": 0.010900243309002433,
+      "grad_norm": 2.983669976646381,
+      "learning_rate": 3.624595469255664e-06,
+      "loss": 0.6187,
+      "step": 112
+    },
+    {
+      "epoch": 0.01099756690997567,
+      "grad_norm": 3.388543821878077,
+      "learning_rate": 3.6569579288025893e-06,
+      "loss": 0.717,
+      "step": 113
+    },
+    {
+      "epoch": 0.011094890510948904,
+      "grad_norm": 3.0720120062792127,
+      "learning_rate": 3.689320388349515e-06,
+      "loss": 0.5057,
+      "step": 114
+    },
+    {
+      "epoch": 0.01119221411192214,
+      "grad_norm": 2.521868238475485,
+      "learning_rate": 3.721682847896441e-06,
+      "loss": 0.4308,
+      "step": 115
+    },
+    {
+      "epoch": 0.011289537712895377,
+      "grad_norm": 2.641085251645149,
+      "learning_rate": 3.754045307443366e-06,
+      "loss": 0.4047,
+      "step": 116
+    },
+    {
+      "epoch": 0.011386861313868613,
+      "grad_norm": 2.6936547530255828,
+      "learning_rate": 3.7864077669902915e-06,
+      "loss": 0.5548,
+      "step": 117
+    },
+    {
+      "epoch": 0.01148418491484185,
+      "grad_norm": 5.599830434139348,
+      "learning_rate": 3.818770226537217e-06,
+      "loss": 0.5338,
+      "step": 118
+    },
+    {
+      "epoch": 0.011581508515815086,
+      "grad_norm": 2.6372065340185378,
+      "learning_rate": 3.851132686084142e-06,
+      "loss": 0.4833,
+      "step": 119
+    },
+    {
+      "epoch": 0.01167883211678832,
+      "grad_norm": 2.555049765563167,
+      "learning_rate": 3.883495145631068e-06,
+      "loss": 0.4295,
+      "step": 120
+    },
+    {
+      "epoch": 0.011776155717761557,
+      "grad_norm": 2.22725048478721,
+      "learning_rate": 3.915857605177994e-06,
+      "loss": 0.4074,
+      "step": 121
+    },
+    {
+      "epoch": 0.011873479318734793,
+      "grad_norm": 3.0093045583939984,
+      "learning_rate": 3.948220064724919e-06,
+      "loss": 0.7168,
+      "step": 122
+    },
+    {
+      "epoch": 0.01197080291970803,
+      "grad_norm": 2.8800338131191223,
+      "learning_rate": 3.980582524271845e-06,
+      "loss": 0.3826,
+      "step": 123
+    },
+    {
+      "epoch": 0.012068126520681266,
+      "grad_norm": 2.3197904571086974,
+      "learning_rate": 4.01294498381877e-06,
+      "loss": 0.2584,
+      "step": 124
+    },
+    {
+      "epoch": 0.012165450121654502,
+      "grad_norm": 2.929540360888414,
+      "learning_rate": 4.045307443365696e-06,
+      "loss": 0.4617,
+      "step": 125
+    },
+    {
+      "epoch": 0.012262773722627737,
+      "grad_norm": 2.5602803735383137,
+      "learning_rate": 4.0776699029126215e-06,
+      "loss": 0.2561,
+      "step": 126
+    },
+    {
+      "epoch": 0.012360097323600973,
+      "grad_norm": 2.676345297957673,
+      "learning_rate": 4.1100323624595475e-06,
+      "loss": 0.2996,
+      "step": 127
+    },
+    {
+      "epoch": 0.01245742092457421,
+      "grad_norm": 1.9047794610871986,
+      "learning_rate": 4.1423948220064734e-06,
+      "loss": 0.3475,
+      "step": 128
+    },
+    {
+      "epoch": 0.012554744525547445,
+      "grad_norm": 2.9014607006450555,
+      "learning_rate": 4.1747572815533986e-06,
+      "loss": 0.4748,
+      "step": 129
+    },
+    {
+      "epoch": 0.012652068126520682,
+      "grad_norm": 2.2992367182815987,
+      "learning_rate": 4.207119741100324e-06,
+      "loss": 0.3465,
+      "step": 130
+    },
+    {
+      "epoch": 0.012749391727493918,
+      "grad_norm": 2.668874383033437,
+      "learning_rate": 4.23948220064725e-06,
+      "loss": 0.6119,
+      "step": 131
+    },
+    {
+      "epoch": 0.012846715328467153,
+      "grad_norm": 2.69106703615133,
+      "learning_rate": 4.271844660194175e-06,
+      "loss": 0.4743,
+      "step": 132
+    },
+    {
+      "epoch": 0.012944038929440389,
+      "grad_norm": 2.972314561813759,
+      "learning_rate": 4.304207119741101e-06,
+      "loss": 0.5766,
+      "step": 133
+    },
+    {
+      "epoch": 0.013041362530413625,
+      "grad_norm": 2.7487017428059635,
+      "learning_rate": 4.336569579288027e-06,
+      "loss": 0.5818,
+      "step": 134
+    },
+    {
+      "epoch": 0.013138686131386862,
+      "grad_norm": 3.1117207482379663,
+      "learning_rate": 4.368932038834952e-06,
+      "loss": 0.6468,
+      "step": 135
+    },
+    {
+      "epoch": 0.013236009732360098,
+      "grad_norm": 2.781796948090657,
+      "learning_rate": 4.401294498381877e-06,
+      "loss": 0.7209,
+      "step": 136
+    },
+    {
+      "epoch": 0.013333333333333334,
+      "grad_norm": 2.5480533986327556,
+      "learning_rate": 4.433656957928803e-06,
+      "loss": 0.5907,
+      "step": 137
+    },
+    {
+      "epoch": 0.013430656934306569,
+      "grad_norm": 2.054397852683208,
+      "learning_rate": 4.466019417475729e-06,
+      "loss": 0.4079,
+      "step": 138
+    },
+    {
+      "epoch": 0.013527980535279805,
+      "grad_norm": 2.2564046621809037,
+      "learning_rate": 4.498381877022654e-06,
+      "loss": 0.4,
+      "step": 139
+    },
+    {
+      "epoch": 0.013625304136253041,
+      "grad_norm": 2.8739841159071022,
+      "learning_rate": 4.53074433656958e-06,
+      "loss": 0.5819,
+      "step": 140
+    },
+    {
+      "epoch": 0.013722627737226278,
+      "grad_norm": 2.6418540847993657,
+      "learning_rate": 4.563106796116505e-06,
+      "loss": 0.589,
+      "step": 141
+    },
+    {
+      "epoch": 0.013819951338199514,
+      "grad_norm": 2.431908870746442,
+      "learning_rate": 4.59546925566343e-06,
+      "loss": 0.5468,
+      "step": 142
+    },
+    {
+      "epoch": 0.013917274939172749,
+      "grad_norm": 4.44933942542394,
+      "learning_rate": 4.627831715210356e-06,
+      "loss": 0.3846,
+      "step": 143
+    },
+    {
+      "epoch": 0.014014598540145985,
+      "grad_norm": 2.2469929628351126,
+      "learning_rate": 4.660194174757282e-06,
+      "loss": 0.3047,
+      "step": 144
+    },
+    {
+      "epoch": 0.014111922141119221,
+      "grad_norm": 2.8361034502388205,
+      "learning_rate": 4.6925566343042074e-06,
+      "loss": 0.4186,
+      "step": 145
+    },
+    {
+      "epoch": 0.014209245742092457,
+      "grad_norm": 2.485184255788147,
+      "learning_rate": 4.724919093851133e-06,
+      "loss": 0.455,
+      "step": 146
+    },
+    {
+      "epoch": 0.014306569343065694,
+      "grad_norm": 2.677307495548506,
+      "learning_rate": 4.7572815533980585e-06,
+      "loss": 0.6346,
+      "step": 147
+    },
+    {
+      "epoch": 0.01440389294403893,
+      "grad_norm": 2.9440091029213034,
+      "learning_rate": 4.789644012944984e-06,
+      "loss": 0.4961,
+      "step": 148
+    },
+    {
+      "epoch": 0.014501216545012165,
+      "grad_norm": 2.6810327828724723,
+      "learning_rate": 4.82200647249191e-06,
+      "loss": 0.3754,
+      "step": 149
+    },
+    {
+      "epoch": 0.014598540145985401,
+      "grad_norm": 2.519257002697837,
+      "learning_rate": 4.854368932038836e-06,
+      "loss": 0.249,
+      "step": 150
+    },
+    {
+      "epoch": 0.014695863746958637,
+      "grad_norm": 2.8041238457488578,
+      "learning_rate": 4.886731391585761e-06,
+      "loss": 0.3117,
+      "step": 151
+    },
+    {
+      "epoch": 0.014793187347931874,
+      "grad_norm": 2.363481194731433,
+      "learning_rate": 4.919093851132687e-06,
+      "loss": 0.3325,
+      "step": 152
+    },
+    {
+      "epoch": 0.01489051094890511,
+      "grad_norm": 3.078347599868747,
+      "learning_rate": 4.951456310679612e-06,
+      "loss": 0.3569,
+      "step": 153
+    },
+    {
+      "epoch": 0.014987834549878346,
+      "grad_norm": 3.2926461094535515,
+      "learning_rate": 4.983818770226538e-06,
+      "loss": 0.716,
+      "step": 154
+    },
+    {
+      "epoch": 0.01508515815085158,
+      "grad_norm": 2.340052421830345,
+      "learning_rate": 5.016181229773464e-06,
+      "loss": 0.2642,
+      "step": 155
+    },
+    {
+      "epoch": 0.015182481751824817,
+      "grad_norm": 1.8915730140906823,
+      "learning_rate": 5.048543689320389e-06,
+      "loss": 0.3523,
+      "step": 156
+    },
+    {
+      "epoch": 0.015279805352798053,
+      "grad_norm": 4.2448533254564484,
+      "learning_rate": 5.080906148867314e-06,
+      "loss": 0.5185,
+      "step": 157
+    },
+    {
+      "epoch": 0.01537712895377129,
+      "grad_norm": 2.1172922256300333,
+      "learning_rate": 5.11326860841424e-06,
+      "loss": 0.3341,
+      "step": 158
+    },
+    {
+      "epoch": 0.015474452554744526,
+      "grad_norm": 2.7414250631657113,
+      "learning_rate": 5.145631067961165e-06,
+      "loss": 0.5965,
+      "step": 159
+    },
+    {
+      "epoch": 0.015571776155717762,
+      "grad_norm": 1.977804344185745,
+      "learning_rate": 5.17799352750809e-06,
+      "loss": 0.239,
+      "step": 160
+    },
+    {
+      "epoch": 0.015669099756690997,
+      "grad_norm": 2.771807640315191,
+      "learning_rate": 5.210355987055017e-06,
+      "loss": 0.4122,
+      "step": 161
+    },
+    {
+      "epoch": 0.015766423357664233,
+      "grad_norm": 1.9977073642008174,
+      "learning_rate": 5.242718446601942e-06,
+      "loss": 0.3423,
+      "step": 162
+    },
+    {
+      "epoch": 0.01586374695863747,
+      "grad_norm": 3.222730527079622,
+      "learning_rate": 5.275080906148867e-06,
+      "loss": 0.5647,
+      "step": 163
+    },
+    {
+      "epoch": 0.015961070559610706,
+      "grad_norm": 2.95441646694508,
+      "learning_rate": 5.307443365695793e-06,
+      "loss": 0.5198,
+      "step": 164
+    },
+    {
+      "epoch": 0.016058394160583942,
+      "grad_norm": 2.3346384576429116,
+      "learning_rate": 5.3398058252427185e-06,
+      "loss": 0.3516,
+      "step": 165
+    },
+    {
+      "epoch": 0.01615571776155718,
+      "grad_norm": 2.089159587923689,
+      "learning_rate": 5.372168284789644e-06,
+      "loss": 0.3704,
+      "step": 166
+    },
+    {
+      "epoch": 0.016253041362530415,
+      "grad_norm": 2.8135820638465088,
+      "learning_rate": 5.4045307443365705e-06,
+      "loss": 0.3729,
+      "step": 167
+    },
+    {
+      "epoch": 0.01635036496350365,
+      "grad_norm": 2.991259557993277,
+      "learning_rate": 5.436893203883496e-06,
+      "loss": 0.5622,
+      "step": 168
+    },
+    {
+      "epoch": 0.016447688564476887,
+      "grad_norm": 3.1512644455187857,
+      "learning_rate": 5.4692556634304216e-06,
+      "loss": 0.5915,
+      "step": 169
+    },
+    {
+      "epoch": 0.01654501216545012,
+      "grad_norm": 2.616126184062516,
+      "learning_rate": 5.501618122977347e-06,
+      "loss": 0.4252,
+      "step": 170
+    },
+    {
+      "epoch": 0.016642335766423356,
+      "grad_norm": 1.9958281517625203,
+      "learning_rate": 5.533980582524272e-06,
+      "loss": 0.3704,
+      "step": 171
+    },
+    {
+      "epoch": 0.016739659367396593,
+      "grad_norm": 2.470731302334384,
+      "learning_rate": 5.566343042071199e-06,
+      "loss": 0.4373,
+      "step": 172
+    },
+    {
+      "epoch": 0.01683698296836983,
+      "grad_norm": 2.583270308023139,
+      "learning_rate": 5.598705501618124e-06,
+      "loss": 0.4125,
+      "step": 173
+    },
+    {
+      "epoch": 0.016934306569343065,
+      "grad_norm": 1.9644684632241667,
+      "learning_rate": 5.631067961165049e-06,
+      "loss": 0.3522,
+      "step": 174
+    },
+    {
+      "epoch": 0.0170316301703163,
+      "grad_norm": 2.4207097357376046,
+      "learning_rate": 5.663430420711975e-06,
+      "loss": 0.3579,
+      "step": 175
+    },
+    {
+      "epoch": 0.017128953771289538,
+      "grad_norm": 2.3511041847292034,
+      "learning_rate": 5.6957928802589e-06,
+      "loss": 0.5412,
+      "step": 176
+    },
+    {
+      "epoch": 0.017226277372262774,
+      "grad_norm": 2.274427899539275,
+      "learning_rate": 5.728155339805825e-06,
+      "loss": 0.5353,
+      "step": 177
+    },
+    {
+      "epoch": 0.01732360097323601,
+      "grad_norm": 2.133749284526256,
+      "learning_rate": 5.760517799352752e-06,
+      "loss": 0.4392,
+      "step": 178
+    },
+    {
+      "epoch": 0.017420924574209247,
+      "grad_norm": 2.3097462109285787,
+      "learning_rate": 5.792880258899677e-06,
+      "loss": 0.4442,
+      "step": 179
+    },
+    {
+      "epoch": 0.017518248175182483,
+      "grad_norm": 2.2128802818602056,
+      "learning_rate": 5.825242718446602e-06,
+      "loss": 0.5635,
+      "step": 180
+    },
+    {
+      "epoch": 0.017615571776155716,
+      "grad_norm": 2.103405792854256,
+      "learning_rate": 5.857605177993528e-06,
+      "loss": 0.4533,
+      "step": 181
+    },
+    {
+      "epoch": 0.017712895377128952,
+      "grad_norm": 2.0565661990183597,
+      "learning_rate": 5.889967637540453e-06,
+      "loss": 0.3806,
+      "step": 182
+    },
+    {
+      "epoch": 0.01781021897810219,
+      "grad_norm": 2.179649872267064,
+      "learning_rate": 5.9223300970873785e-06,
+      "loss": 0.3842,
+      "step": 183
+    },
+    {
+      "epoch": 0.017907542579075425,
+      "grad_norm": 3.8333244047199146,
+      "learning_rate": 5.954692556634305e-06,
+      "loss": 0.3876,
+      "step": 184
+    },
+    {
+      "epoch": 0.01800486618004866,
+      "grad_norm": 2.2893517217095716,
+      "learning_rate": 5.9870550161812304e-06,
+      "loss": 0.4781,
+      "step": 185
+    },
+    {
+      "epoch": 0.018102189781021898,
+      "grad_norm": 1.6022498167897639,
+      "learning_rate": 6.0194174757281556e-06,
+      "loss": 0.2306,
+      "step": 186
+    },
+    {
+      "epoch": 0.018199513381995134,
+      "grad_norm": 2.32863493589546,
+      "learning_rate": 6.0517799352750815e-06,
+      "loss": 0.5139,
+      "step": 187
+    },
+    {
+      "epoch": 0.01829683698296837,
+      "grad_norm": 2.0789478938631314,
+      "learning_rate": 6.084142394822007e-06,
+      "loss": 0.2824,
+      "step": 188
+    },
+    {
+      "epoch": 0.018394160583941607,
+      "grad_norm": 1.7544615955949223,
+      "learning_rate": 6.116504854368932e-06,
+      "loss": 0.4172,
+      "step": 189
+    },
+    {
+      "epoch": 0.018491484184914843,
+      "grad_norm": 1.931043696572374,
+      "learning_rate": 6.148867313915859e-06,
+      "loss": 0.3584,
+      "step": 190
+    },
+    {
+      "epoch": 0.01858880778588808,
+      "grad_norm": 2.467258437370788,
+      "learning_rate": 6.181229773462784e-06,
+      "loss": 0.462,
+      "step": 191
+    },
+    {
+      "epoch": 0.018686131386861315,
+      "grad_norm": 2.1541091684996965,
+      "learning_rate": 6.213592233009709e-06,
+      "loss": 0.3967,
+      "step": 192
+    },
+    {
+      "epoch": 0.01878345498783455,
+      "grad_norm": 2.2330486922808395,
+      "learning_rate": 6.245954692556635e-06,
+      "loss": 0.5316,
+      "step": 193
+    },
+    {
+      "epoch": 0.018880778588807785,
+      "grad_norm": 2.3498262097642395,
+      "learning_rate": 6.27831715210356e-06,
+      "loss": 0.4815,
+      "step": 194
+    },
+    {
+      "epoch": 0.01897810218978102,
+      "grad_norm": 1.7045092076002246,
+      "learning_rate": 6.310679611650487e-06,
+      "loss": 0.3,
+      "step": 195
+    },
+    {
+      "epoch": 0.019075425790754257,
+      "grad_norm": 2.5703331850837023,
+      "learning_rate": 6.343042071197412e-06,
+      "loss": 0.4143,
+      "step": 196
+    },
+    {
+      "epoch": 0.019172749391727494,
+      "grad_norm": 2.6940646171495133,
+      "learning_rate": 6.375404530744337e-06,
+      "loss": 0.5463,
+      "step": 197
+    },
+    {
+      "epoch": 0.01927007299270073,
+      "grad_norm": 2.4185580273524847,
+      "learning_rate": 6.407766990291263e-06,
+      "loss": 0.5215,
+      "step": 198
+    },
+    {
+      "epoch": 0.019367396593673966,
+      "grad_norm": 2.6509824694985946,
+      "learning_rate": 6.440129449838188e-06,
+      "loss": 0.5286,
+      "step": 199
+    },
+    {
+      "epoch": 0.019464720194647202,
+      "grad_norm": 2.4807219128312767,
+      "learning_rate": 6.472491909385113e-06,
+      "loss": 0.3996,
+      "step": 200
+    },
+    {
+      "epoch": 0.01956204379562044,
+      "grad_norm": 2.651883834043772,
+      "learning_rate": 6.50485436893204e-06,
+      "loss": 0.3499,
+      "step": 201
+    },
+    {
+      "epoch": 0.019659367396593675,
+      "grad_norm": 2.670759179984812,
+      "learning_rate": 6.537216828478965e-06,
+      "loss": 0.552,
+      "step": 202
+    },
+    {
+      "epoch": 0.01975669099756691,
+      "grad_norm": 2.51305850829245,
+      "learning_rate": 6.56957928802589e-06,
+      "loss": 0.3806,
+      "step": 203
+    },
+    {
+      "epoch": 0.019854014598540148,
+      "grad_norm": 2.435954851305265,
+      "learning_rate": 6.601941747572816e-06,
+      "loss": 0.6093,
+      "step": 204
+    },
+    {
+      "epoch": 0.01995133819951338,
+      "grad_norm": 2.091315833022872,
+      "learning_rate": 6.6343042071197415e-06,
+      "loss": 0.3573,
+      "step": 205
+    },
+    {
+      "epoch": 0.020048661800486617,
+      "grad_norm": 2.205515437184344,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.2892,
+      "step": 206
+    },
+    {
+      "epoch": 0.020145985401459853,
+      "grad_norm": 2.314981932930035,
+      "learning_rate": 6.6990291262135935e-06,
+      "loss": 0.4184,
+      "step": 207
+    },
+    {
+      "epoch": 0.02024330900243309,
+      "grad_norm": 1.9102474885146974,
+      "learning_rate": 6.731391585760519e-06,
+      "loss": 0.2287,
+      "step": 208
+    },
+    {
+      "epoch": 0.020340632603406326,
+      "grad_norm": 1.9408029275065433,
+      "learning_rate": 6.763754045307444e-06,
+      "loss": 0.3958,
+      "step": 209
+    },
+    {
+      "epoch": 0.020437956204379562,
+      "grad_norm": 2.1006467731485823,
+      "learning_rate": 6.79611650485437e-06,
+      "loss": 0.3764,
+      "step": 210
+    },
+    {
+      "epoch": 0.0205352798053528,
+      "grad_norm": 2.0927447282795146,
+      "learning_rate": 6.828478964401295e-06,
+      "loss": 0.531,
+      "step": 211
+    },
+    {
+      "epoch": 0.020632603406326035,
+      "grad_norm": 3.4830081465453633,
+      "learning_rate": 6.86084142394822e-06,
+      "loss": 0.4887,
+      "step": 212
+    },
+    {
+      "epoch": 0.02072992700729927,
+      "grad_norm": 2.253360993066953,
+      "learning_rate": 6.893203883495147e-06,
+      "loss": 0.4587,
+      "step": 213
+    },
+    {
+      "epoch": 0.020827250608272507,
+      "grad_norm": 3.3751096354443852,
+      "learning_rate": 6.925566343042072e-06,
+      "loss": 0.3427,
+      "step": 214
+    },
+    {
+      "epoch": 0.020924574209245744,
+      "grad_norm": 1.9729713112803993,
+      "learning_rate": 6.957928802588997e-06,
+      "loss": 0.384,
+      "step": 215
+    },
+    {
+      "epoch": 0.021021897810218976,
+      "grad_norm": 2.761285167796522,
+      "learning_rate": 6.990291262135923e-06,
+      "loss": 0.3512,
+      "step": 216
+    },
+    {
+      "epoch": 0.021119221411192213,
+      "grad_norm": 2.431882400442612,
+      "learning_rate": 7.022653721682848e-06,
+      "loss": 0.3971,
+      "step": 217
+    },
+    {
+      "epoch": 0.02121654501216545,
+      "grad_norm": 3.659254877088116,
+      "learning_rate": 7.055016181229773e-06,
+      "loss": 0.4115,
+      "step": 218
+    },
+    {
+      "epoch": 0.021313868613138685,
+      "grad_norm": 2.5501534359714655,
+      "learning_rate": 7.0873786407767e-06,
+      "loss": 0.4963,
+      "step": 219
+    },
+    {
+      "epoch": 0.02141119221411192,
+      "grad_norm": 4.296894309260591,
+      "learning_rate": 7.119741100323625e-06,
+      "loss": 0.5203,
+      "step": 220
+    },
+    {
+      "epoch": 0.021508515815085158,
+      "grad_norm": 2.5489854552137237,
+      "learning_rate": 7.152103559870551e-06,
+      "loss": 0.4343,
+      "step": 221
+    },
+    {
+      "epoch": 0.021605839416058394,
+      "grad_norm": 2.00955207958064,
+      "learning_rate": 7.184466019417476e-06,
+      "loss": 0.3603,
+      "step": 222
+    },
+    {
+      "epoch": 0.02170316301703163,
+      "grad_norm": 2.2675038932590224,
+      "learning_rate": 7.2168284789644015e-06,
+      "loss": 0.3968,
+      "step": 223
+    },
+    {
+      "epoch": 0.021800486618004867,
+      "grad_norm": 2.4690586331753277,
+      "learning_rate": 7.249190938511328e-06,
+      "loss": 0.5883,
+      "step": 224
+    },
+    {
+      "epoch": 0.021897810218978103,
+      "grad_norm": 2.141328682063472,
+      "learning_rate": 7.2815533980582534e-06,
+      "loss": 0.3547,
+      "step": 225
+    },
+    {
+      "epoch": 0.02199513381995134,
+      "grad_norm": 2.223927434368622,
+      "learning_rate": 7.3139158576051786e-06,
+      "loss": 0.5031,
+      "step": 226
+    },
+    {
+      "epoch": 0.022092457420924576,
+      "grad_norm": 2.8602320319532346,
+      "learning_rate": 7.3462783171521046e-06,
+      "loss": 0.4226,
+      "step": 227
+    },
+    {
+      "epoch": 0.02218978102189781,
+      "grad_norm": 2.8852449405031835,
+      "learning_rate": 7.37864077669903e-06,
+      "loss": 0.4298,
+      "step": 228
+    },
+    {
+      "epoch": 0.022287104622871045,
+      "grad_norm": 1.7370522944561966,
+      "learning_rate": 7.411003236245955e-06,
+      "loss": 0.3827,
+      "step": 229
+    },
+    {
+      "epoch": 0.02238442822384428,
+      "grad_norm": 2.3907908463140584,
+      "learning_rate": 7.443365695792882e-06,
+      "loss": 0.4139,
+      "step": 230
+    },
+    {
+      "epoch": 0.022481751824817518,
+      "grad_norm": 2.27581306432663,
+      "learning_rate": 7.475728155339807e-06,
+      "loss": 0.4736,
+      "step": 231
+    },
+    {
+      "epoch": 0.022579075425790754,
+      "grad_norm": 2.1861094823645675,
+      "learning_rate": 7.508090614886732e-06,
+      "loss": 0.4809,
+      "step": 232
+    },
+    {
+      "epoch": 0.02267639902676399,
+      "grad_norm": 1.9626208371421419,
+      "learning_rate": 7.540453074433658e-06,
+      "loss": 0.3436,
+      "step": 233
+    },
+    {
+      "epoch": 0.022773722627737226,
+      "grad_norm": 1.7092390993202267,
+      "learning_rate": 7.572815533980583e-06,
+      "loss": 0.3224,
+      "step": 234
+    },
+    {
+      "epoch": 0.022871046228710463,
+      "grad_norm": 3.0168693228526546,
+      "learning_rate": 7.605177993527508e-06,
+      "loss": 0.6366,
+      "step": 235
+    },
+    {
+      "epoch": 0.0229683698296837,
+      "grad_norm": 2.424919921496664,
+      "learning_rate": 7.637540453074434e-06,
+      "loss": 0.4483,
+      "step": 236
+    },
+    {
+      "epoch": 0.023065693430656935,
+      "grad_norm": 2.4586833984787626,
+      "learning_rate": 7.66990291262136e-06,
+      "loss": 0.4031,
+      "step": 237
+    },
+    {
+      "epoch": 0.02316301703163017,
+      "grad_norm": 2.092010230715883,
+      "learning_rate": 7.702265372168284e-06,
+      "loss": 0.4257,
+      "step": 238
+    },
+    {
+      "epoch": 0.023260340632603408,
+      "grad_norm": 2.3360188447701655,
+      "learning_rate": 7.734627831715211e-06,
+      "loss": 0.4684,
+      "step": 239
+    },
+    {
+      "epoch": 0.02335766423357664,
+      "grad_norm": 2.087175894606599,
+      "learning_rate": 7.766990291262136e-06,
+      "loss": 0.4272,
+      "step": 240
+    },
+    {
+      "epoch": 0.023454987834549877,
+      "grad_norm": 2.598684557686617,
+      "learning_rate": 7.799352750809061e-06,
+      "loss": 0.5401,
+      "step": 241
+    },
+    {
+      "epoch": 0.023552311435523113,
+      "grad_norm": 2.025117037181364,
+      "learning_rate": 7.831715210355988e-06,
+      "loss": 0.372,
+      "step": 242
+    },
+    {
+      "epoch": 0.02364963503649635,
+      "grad_norm": 2.2467324584398405,
+      "learning_rate": 7.864077669902913e-06,
+      "loss": 0.5891,
+      "step": 243
+    },
+    {
+      "epoch": 0.023746958637469586,
+      "grad_norm": 2.38036373195977,
+      "learning_rate": 7.896440129449839e-06,
+      "loss": 0.5133,
+      "step": 244
+    },
+    {
+      "epoch": 0.023844282238442822,
+      "grad_norm": 2.052700924442009,
+      "learning_rate": 7.928802588996765e-06,
+      "loss": 0.5161,
+      "step": 245
+    },
+    {
+      "epoch": 0.02394160583941606,
+      "grad_norm": 3.4299018810240254,
+      "learning_rate": 7.96116504854369e-06,
+      "loss": 0.5314,
+      "step": 246
+    },
+    {
+      "epoch": 0.024038929440389295,
+      "grad_norm": 1.3903956706369247,
+      "learning_rate": 7.993527508090616e-06,
+      "loss": 0.3539,
+      "step": 247
+    },
+    {
+      "epoch": 0.02413625304136253,
+      "grad_norm": 2.4599878810180873,
+      "learning_rate": 8.02588996763754e-06,
+      "loss": 0.4876,
+      "step": 248
+    },
+    {
+      "epoch": 0.024233576642335768,
+      "grad_norm": 2.4053308291912083,
+      "learning_rate": 8.058252427184466e-06,
+      "loss": 0.5185,
+      "step": 249
+    },
+    {
+      "epoch": 0.024330900243309004,
+      "grad_norm": 1.6624263546342495,
+      "learning_rate": 8.090614886731393e-06,
+      "loss": 0.2909,
+      "step": 250
+    },
+    {
+      "epoch": 0.024428223844282237,
+      "grad_norm": 2.4091367373679597,
+      "learning_rate": 8.122977346278318e-06,
+      "loss": 0.6192,
+      "step": 251
+    },
+    {
+      "epoch": 0.024525547445255473,
+      "grad_norm": 2.4595313520548427,
+      "learning_rate": 8.155339805825243e-06,
+      "loss": 0.3444,
+      "step": 252
+    },
+    {
+      "epoch": 0.02462287104622871,
+      "grad_norm": 2.3200411140153174,
+      "learning_rate": 8.18770226537217e-06,
+      "loss": 0.6112,
+      "step": 253
+    },
+    {
+      "epoch": 0.024720194647201946,
+      "grad_norm": 2.029624875741936,
+      "learning_rate": 8.220064724919095e-06,
+      "loss": 0.4524,
+      "step": 254
+    },
+    {
+      "epoch": 0.024817518248175182,
+      "grad_norm": 1.8862765408033388,
+      "learning_rate": 8.25242718446602e-06,
+      "loss": 0.2173,
+      "step": 255
+    },
+    {
+      "epoch": 0.02491484184914842,
+      "grad_norm": 2.575687620331568,
+      "learning_rate": 8.284789644012947e-06,
+      "loss": 0.4599,
+      "step": 256
+    },
+    {
+      "epoch": 0.025012165450121655,
+      "grad_norm": 2.373530485379713,
+      "learning_rate": 8.317152103559872e-06,
+      "loss": 0.5326,
+      "step": 257
+    },
+    {
+      "epoch": 0.02510948905109489,
+      "grad_norm": 2.4086353319447262,
+      "learning_rate": 8.349514563106797e-06,
+      "loss": 0.6275,
+      "step": 258
+    },
+    {
+      "epoch": 0.025206812652068127,
+      "grad_norm": 2.1075725625285697,
+      "learning_rate": 8.381877022653722e-06,
+      "loss": 0.44,
+      "step": 259
+    },
+    {
+      "epoch": 0.025304136253041364,
+      "grad_norm": 2.0285700798989614,
+      "learning_rate": 8.414239482200647e-06,
+      "loss": 0.3489,
+      "step": 260
+    },
+    {
+      "epoch": 0.0254014598540146,
+      "grad_norm": 2.5592973746241,
+      "learning_rate": 8.446601941747573e-06,
+      "loss": 0.4403,
+      "step": 261
+    },
+    {
+      "epoch": 0.025498783454987836,
+      "grad_norm": 2.470930078509074,
+      "learning_rate": 8.4789644012945e-06,
+      "loss": 0.4985,
+      "step": 262
+    },
+    {
+      "epoch": 0.02559610705596107,
+      "grad_norm": 2.099638103909556,
+      "learning_rate": 8.511326860841424e-06,
+      "loss": 0.4194,
+      "step": 263
+    },
+    {
+      "epoch": 0.025693430656934305,
+      "grad_norm": 1.6030834140551835,
+      "learning_rate": 8.54368932038835e-06,
+      "loss": 0.3382,
+      "step": 264
+    },
+    {
+      "epoch": 0.02579075425790754,
+      "grad_norm": 1.8960928547169034,
+      "learning_rate": 8.576051779935276e-06,
+      "loss": 0.2838,
+      "step": 265
+    },
+    {
+      "epoch": 0.025888077858880778,
+      "grad_norm": 2.4306930963261966,
+      "learning_rate": 8.608414239482202e-06,
+      "loss": 0.4956,
+      "step": 266
+    },
+    {
+      "epoch": 0.025985401459854014,
+      "grad_norm": 2.374430136325354,
+      "learning_rate": 8.640776699029127e-06,
+      "loss": 0.5083,
+      "step": 267
+    },
+    {
+      "epoch": 0.02608272506082725,
+      "grad_norm": 2.410095115145934,
+      "learning_rate": 8.673139158576054e-06,
+      "loss": 0.4247,
+      "step": 268
+    },
+    {
+      "epoch": 0.026180048661800487,
+      "grad_norm": 2.41271065696519,
+      "learning_rate": 8.705501618122979e-06,
+      "loss": 0.6946,
+      "step": 269
+    },
+    {
+      "epoch": 0.026277372262773723,
+      "grad_norm": 1.752688930628829,
+      "learning_rate": 8.737864077669904e-06,
+      "loss": 0.2662,
+      "step": 270
+    },
+    {
+      "epoch": 0.02637469586374696,
+      "grad_norm": 1.9842034213162434,
+      "learning_rate": 8.770226537216829e-06,
+      "loss": 0.3611,
+      "step": 271
+    },
+    {
+      "epoch": 0.026472019464720196,
+      "grad_norm": 2.4137979998327497,
+      "learning_rate": 8.802588996763754e-06,
+      "loss": 0.501,
+      "step": 272
+    },
+    {
+      "epoch": 0.026569343065693432,
+      "grad_norm": 2.929650064864996,
+      "learning_rate": 8.834951456310681e-06,
+      "loss": 0.6153,
+      "step": 273
+    },
+    {
+      "epoch": 0.02666666666666667,
+      "grad_norm": 2.281738020025263,
+      "learning_rate": 8.867313915857606e-06,
+      "loss": 0.5395,
+      "step": 274
+    },
+    {
+      "epoch": 0.0267639902676399,
+      "grad_norm": 2.1406726692627975,
+      "learning_rate": 8.899676375404531e-06,
+      "loss": 0.4039,
+      "step": 275
+    },
+    {
+      "epoch": 0.026861313868613138,
+      "grad_norm": 3.2366954201371523,
+      "learning_rate": 8.932038834951458e-06,
+      "loss": 0.5414,
+      "step": 276
+    },
+    {
+      "epoch": 0.026958637469586374,
+      "grad_norm": 2.1900667662872513,
+      "learning_rate": 8.964401294498383e-06,
+      "loss": 0.3815,
+      "step": 277
+    },
+    {
+      "epoch": 0.02705596107055961,
+      "grad_norm": 2.5301939091612216,
+      "learning_rate": 8.996763754045308e-06,
+      "loss": 0.8016,
+      "step": 278
+    },
+    {
+      "epoch": 0.027153284671532846,
+      "grad_norm": 2.2552758985680907,
+      "learning_rate": 9.029126213592233e-06,
+      "loss": 0.4133,
+      "step": 279
+    },
+    {
+      "epoch": 0.027250608272506083,
+      "grad_norm": 2.309545536997134,
+      "learning_rate": 9.06148867313916e-06,
+      "loss": 0.5346,
+      "step": 280
+    },
+    {
+      "epoch": 0.02734793187347932,
+      "grad_norm": 2.585578916644781,
+      "learning_rate": 9.093851132686085e-06,
+      "loss": 0.407,
+      "step": 281
+    },
+    {
+      "epoch": 0.027445255474452555,
+      "grad_norm": 1.8503464194025006,
+      "learning_rate": 9.12621359223301e-06,
+      "loss": 0.4674,
+      "step": 282
+    },
+    {
+      "epoch": 0.02754257907542579,
+      "grad_norm": 2.431490115980846,
+      "learning_rate": 9.158576051779936e-06,
+      "loss": 0.6026,
+      "step": 283
+    },
+    {
+      "epoch": 0.027639902676399028,
+      "grad_norm": 1.916233248702735,
+      "learning_rate": 9.19093851132686e-06,
+      "loss": 0.4949,
+      "step": 284
+    },
+    {
+      "epoch": 0.027737226277372264,
+      "grad_norm": 2.2160236640245072,
+      "learning_rate": 9.223300970873788e-06,
+      "loss": 0.4765,
+      "step": 285
+    },
+    {
+      "epoch": 0.027834549878345497,
+      "grad_norm": 2.0764827118780143,
+      "learning_rate": 9.255663430420713e-06,
+      "loss": 0.472,
+      "step": 286
+    },
+    {
+      "epoch": 0.027931873479318733,
+      "grad_norm": 2.638286661284288,
+      "learning_rate": 9.288025889967638e-06,
+      "loss": 0.6312,
+      "step": 287
+    },
+    {
+      "epoch": 0.02802919708029197,
+      "grad_norm": 1.940011273577467,
+      "learning_rate": 9.320388349514565e-06,
+      "loss": 0.4555,
+      "step": 288
+    },
+    {
+      "epoch": 0.028126520681265206,
+      "grad_norm": 1.8760624736314784,
+      "learning_rate": 9.35275080906149e-06,
+      "loss": 0.3625,
+      "step": 289
+    },
+    {
+      "epoch": 0.028223844282238442,
+      "grad_norm": 1.3468692859077058,
+      "learning_rate": 9.385113268608415e-06,
+      "loss": 0.2442,
+      "step": 290
+    },
+    {
+      "epoch": 0.02832116788321168,
+      "grad_norm": 2.1497394847504014,
+      "learning_rate": 9.41747572815534e-06,
+      "loss": 0.5227,
+      "step": 291
+    },
+    {
+      "epoch": 0.028418491484184915,
+      "grad_norm": 2.1233743171190014,
+      "learning_rate": 9.449838187702267e-06,
+      "loss": 0.6184,
+      "step": 292
+    },
+    {
+      "epoch": 0.02851581508515815,
+      "grad_norm": 2.337806183860394,
+      "learning_rate": 9.482200647249192e-06,
+      "loss": 0.5491,
+      "step": 293
+    },
+    {
+      "epoch": 0.028613138686131388,
+      "grad_norm": 2.015000594070385,
+      "learning_rate": 9.514563106796117e-06,
+      "loss": 0.5137,
+      "step": 294
+    },
+    {
+      "epoch": 0.028710462287104624,
+      "grad_norm": 2.0267324830753766,
+      "learning_rate": 9.546925566343042e-06,
+      "loss": 0.4117,
+      "step": 295
+    },
+    {
+      "epoch": 0.02880778588807786,
+      "grad_norm": 1.732639028192012,
+      "learning_rate": 9.579288025889967e-06,
+      "loss": 0.3156,
+      "step": 296
+    },
+    {
+      "epoch": 0.028905109489051097,
+      "grad_norm": 2.1204056159243923,
+      "learning_rate": 9.611650485436894e-06,
+      "loss": 0.6056,
+      "step": 297
+    },
+    {
+      "epoch": 0.02900243309002433,
+      "grad_norm": 1.7868071753968195,
+      "learning_rate": 9.64401294498382e-06,
+      "loss": 0.3417,
+      "step": 298
+    },
+    {
+      "epoch": 0.029099756690997566,
+      "grad_norm": 1.9477439300595292,
+      "learning_rate": 9.676375404530746e-06,
+      "loss": 0.3631,
+      "step": 299
+    },
+    {
+      "epoch": 0.029197080291970802,
+      "grad_norm": 1.7688147839655162,
+      "learning_rate": 9.708737864077671e-06,
+      "loss": 0.3605,
+      "step": 300
+    },
+    {
+      "epoch": 0.029294403892944038,
+      "grad_norm": 1.9162335597538034,
+      "learning_rate": 9.741100323624596e-06,
+      "loss": 0.2498,
+      "step": 301
+    },
+    {
+      "epoch": 0.029391727493917275,
+      "grad_norm": 2.9282579520055756,
+      "learning_rate": 9.773462783171522e-06,
+      "loss": 0.4286,
+      "step": 302
+    },
+    {
+      "epoch": 0.02948905109489051,
+      "grad_norm": 1.9744499285549086,
+      "learning_rate": 9.805825242718447e-06,
+      "loss": 0.3391,
+      "step": 303
+    },
+    {
+      "epoch": 0.029586374695863747,
+      "grad_norm": 2.2116032868392455,
+      "learning_rate": 9.838187702265373e-06,
+      "loss": 0.3414,
+      "step": 304
+    },
+    {
+      "epoch": 0.029683698296836983,
+      "grad_norm": 1.9159144570242486,
+      "learning_rate": 9.870550161812299e-06,
+      "loss": 0.2915,
+      "step": 305
+    },
+    {
+      "epoch": 0.02978102189781022,
+      "grad_norm": 2.671718838238437,
+      "learning_rate": 9.902912621359224e-06,
+      "loss": 0.79,
+      "step": 306
+    },
+    {
+      "epoch": 0.029878345498783456,
+      "grad_norm": 2.093937424199301,
+      "learning_rate": 9.935275080906149e-06,
+      "loss": 0.576,
+      "step": 307
+    },
+    {
+      "epoch": 0.029975669099756692,
+      "grad_norm": 1.895574286512308,
+      "learning_rate": 9.967637540453076e-06,
+      "loss": 0.4223,
+      "step": 308
+    },
+    {
+      "epoch": 0.03007299270072993,
+      "grad_norm": 2.142643554578675,
+      "learning_rate": 1e-05,
+      "loss": 0.4719,
+      "step": 309
+    },
+    {
+      "epoch": 0.03017031630170316,
+      "grad_norm": 2.2548613483238378,
+      "learning_rate": 9.999999751573464e-06,
+      "loss": 0.5547,
+      "step": 310
+    },
+    {
+      "epoch": 0.030267639902676398,
+      "grad_norm": 2.375146158639999,
+      "learning_rate": 9.99999900629388e-06,
+      "loss": 0.3864,
+      "step": 311
+    },
+    {
+      "epoch": 0.030364963503649634,
+      "grad_norm": 1.558937895217452,
+      "learning_rate": 9.99999776416132e-06,
+      "loss": 0.3409,
+      "step": 312
+    },
+    {
+      "epoch": 0.03046228710462287,
+      "grad_norm": 2.7508940543848115,
+      "learning_rate": 9.99999602517591e-06,
+      "loss": 0.3641,
+      "step": 313
+    },
+    {
+      "epoch": 0.030559610705596107,
+      "grad_norm": 2.228096737889712,
+      "learning_rate": 9.99999378933782e-06,
+      "loss": 0.6464,
+      "step": 314
+    },
+    {
+      "epoch": 0.030656934306569343,
+      "grad_norm": 1.5612763763472,
+      "learning_rate": 9.999991056647274e-06,
+      "loss": 0.3124,
+      "step": 315
+    },
+    {
+      "epoch": 0.03075425790754258,
+      "grad_norm": 2.3203527787434104,
+      "learning_rate": 9.999987827104544e-06,
+      "loss": 0.5893,
+      "step": 316
+    },
+    {
+      "epoch": 0.030851581508515816,
+      "grad_norm": 1.8472611567410342,
+      "learning_rate": 9.999984100709951e-06,
+      "loss": 0.3732,
+      "step": 317
+    },
+    {
+      "epoch": 0.030948905109489052,
+      "grad_norm": 2.269778108549014,
+      "learning_rate": 9.999979877463866e-06,
+      "loss": 0.5537,
+      "step": 318
+    },
+    {
+      "epoch": 0.03104622871046229,
+      "grad_norm": 2.381498581134022,
+      "learning_rate": 9.999975157366705e-06,
+      "loss": 0.7179,
+      "step": 319
+    },
+    {
+      "epoch": 0.031143552311435525,
+      "grad_norm": 1.7030655036823346,
+      "learning_rate": 9.99996994041894e-06,
+      "loss": 0.4256,
+      "step": 320
+    },
+    {
+      "epoch": 0.031240875912408757,
+      "grad_norm": 1.8361141038730153,
+      "learning_rate": 9.999964226621089e-06,
+      "loss": 0.4648,
+      "step": 321
+    },
+    {
+      "epoch": 0.031338199513381994,
+      "grad_norm": 1.7985459229558753,
+      "learning_rate": 9.99995801597372e-06,
+      "loss": 0.3031,
+      "step": 322
+    },
+    {
+      "epoch": 0.031435523114355234,
+      "grad_norm": 2.4309020119442915,
+      "learning_rate": 9.99995130847745e-06,
+      "loss": 0.5011,
+      "step": 323
+    },
+    {
+      "epoch": 0.031532846715328466,
+      "grad_norm": 2.048514022969095,
+      "learning_rate": 9.999944104132944e-06,
+      "loss": 0.6152,
+      "step": 324
+    },
+    {
+      "epoch": 0.031630170316301706,
+      "grad_norm": 1.8892667320795724,
+      "learning_rate": 9.99993640294092e-06,
+      "loss": 0.4738,
+      "step": 325
+    },
+    {
+      "epoch": 0.03172749391727494,
+      "grad_norm": 2.081179331819785,
+      "learning_rate": 9.999928204902141e-06,
+      "loss": 0.5192,
+      "step": 326
+    },
+    {
+      "epoch": 0.03182481751824817,
+      "grad_norm": 2.410280889073595,
+      "learning_rate": 9.999919510017424e-06,
+      "loss": 0.3314,
+      "step": 327
+    },
+    {
+      "epoch": 0.03192214111922141,
+      "grad_norm": 1.663034255724975,
+      "learning_rate": 9.999910318287632e-06,
+      "loss": 0.3342,
+      "step": 328
+    },
+    {
+      "epoch": 0.032019464720194644,
+      "grad_norm": 1.7874391345352068,
+      "learning_rate": 9.999900629713679e-06,
+      "loss": 0.3189,
+      "step": 329
+    },
+    {
+      "epoch": 0.032116788321167884,
+      "grad_norm": 2.1098429973097805,
+      "learning_rate": 9.999890444296528e-06,
+      "loss": 0.4561,
+      "step": 330
+    },
+    {
+      "epoch": 0.03221411192214112,
+      "grad_norm": 2.4678279265353558,
+      "learning_rate": 9.999879762037187e-06,
+      "loss": 0.5831,
+      "step": 331
+    },
+    {
+      "epoch": 0.03231143552311436,
+      "grad_norm": 1.6643716587630457,
+      "learning_rate": 9.999868582936726e-06,
+      "loss": 0.4371,
+      "step": 332
+    },
+    {
+      "epoch": 0.03240875912408759,
+      "grad_norm": 2.088466639768523,
+      "learning_rate": 9.999856906996246e-06,
+      "loss": 0.3904,
+      "step": 333
+    },
+    {
+      "epoch": 0.03250608272506083,
+      "grad_norm": 2.0023651443392256,
+      "learning_rate": 9.999844734216914e-06,
+      "loss": 0.4802,
+      "step": 334
+    },
+    {
+      "epoch": 0.03260340632603406,
+      "grad_norm": 2.161282844007076,
+      "learning_rate": 9.99983206459994e-06,
+      "loss": 0.5187,
+      "step": 335
+    },
+    {
+      "epoch": 0.0327007299270073,
+      "grad_norm": 2.10212671583593,
+      "learning_rate": 9.999818898146576e-06,
+      "loss": 0.4618,
+      "step": 336
+    },
+    {
+      "epoch": 0.032798053527980535,
+      "grad_norm": 2.2142899508809286,
+      "learning_rate": 9.999805234858137e-06,
+      "loss": 0.2387,
+      "step": 337
+    },
+    {
+      "epoch": 0.032895377128953775,
+      "grad_norm": 2.1084763484693023,
+      "learning_rate": 9.999791074735981e-06,
+      "loss": 0.5652,
+      "step": 338
+    },
+    {
+      "epoch": 0.03299270072992701,
+      "grad_norm": 2.261838498017328,
+      "learning_rate": 9.99977641778151e-06,
+      "loss": 0.7224,
+      "step": 339
+    },
+    {
+      "epoch": 0.03309002433090024,
+      "grad_norm": 1.612816030006559,
+      "learning_rate": 9.999761263996184e-06,
+      "loss": 0.377,
+      "step": 340
+    },
+    {
+      "epoch": 0.03318734793187348,
+      "grad_norm": 2.1209830295615832,
+      "learning_rate": 9.999745613381507e-06,
+      "loss": 0.614,
+      "step": 341
+    },
+    {
+      "epoch": 0.03328467153284671,
+      "grad_norm": 1.7938764015879674,
+      "learning_rate": 9.999729465939036e-06,
+      "loss": 0.3983,
+      "step": 342
+    },
+    {
+      "epoch": 0.03338199513381995,
+      "grad_norm": 1.943418875698731,
+      "learning_rate": 9.999712821670375e-06,
+      "loss": 0.4708,
+      "step": 343
+    },
+    {
+      "epoch": 0.033479318734793186,
+      "grad_norm": 1.9787546900237571,
+      "learning_rate": 9.99969568057718e-06,
+      "loss": 0.578,
+      "step": 344
+    },
+    {
+      "epoch": 0.033576642335766425,
+      "grad_norm": 1.4798263726328331,
+      "learning_rate": 9.99967804266115e-06,
+      "loss": 0.394,
+      "step": 345
+    },
+    {
+      "epoch": 0.03367396593673966,
+      "grad_norm": 2.1936298043995484,
+      "learning_rate": 9.99965990792404e-06,
+      "loss": 0.6316,
+      "step": 346
+    },
+    {
+      "epoch": 0.0337712895377129,
+      "grad_norm": 2.2799650780195133,
+      "learning_rate": 9.99964127636765e-06,
+      "loss": 0.3985,
+      "step": 347
+    },
+    {
+      "epoch": 0.03386861313868613,
+      "grad_norm": 1.8519049219191819,
+      "learning_rate": 9.999622147993837e-06,
+      "loss": 0.3853,
+      "step": 348
+    },
+    {
+      "epoch": 0.03396593673965937,
+      "grad_norm": 1.5111895282974241,
+      "learning_rate": 9.999602522804497e-06,
+      "loss": 0.4201,
+      "step": 349
+    },
+    {
+      "epoch": 0.0340632603406326,
+      "grad_norm": 1.8605769784283237,
+      "learning_rate": 9.99958240080158e-06,
+      "loss": 0.5225,
+      "step": 350
+    },
+    {
+      "epoch": 0.034160583941605836,
+      "grad_norm": 1.6063240538866903,
+      "learning_rate": 9.999561781987087e-06,
+      "loss": 0.3165,
+      "step": 351
+    },
+    {
+      "epoch": 0.034257907542579076,
+      "grad_norm": 1.4751976204077173,
+      "learning_rate": 9.999540666363068e-06,
+      "loss": 0.3156,
+      "step": 352
+    },
+    {
+      "epoch": 0.03435523114355231,
+      "grad_norm": 2.1029966771511757,
+      "learning_rate": 9.99951905393162e-06,
+      "loss": 0.5336,
+      "step": 353
+    },
+    {
+      "epoch": 0.03445255474452555,
+      "grad_norm": 2.1419054642874267,
+      "learning_rate": 9.99949694469489e-06,
+      "loss": 0.5253,
+      "step": 354
+    },
+    {
+      "epoch": 0.03454987834549878,
+      "grad_norm": 2.169397271826959,
+      "learning_rate": 9.999474338655075e-06,
+      "loss": 0.5567,
+      "step": 355
+    },
+    {
+      "epoch": 0.03464720194647202,
+      "grad_norm": 2.2972412855327797,
+      "learning_rate": 9.999451235814422e-06,
+      "loss": 0.5233,
+      "step": 356
+    },
+    {
+      "epoch": 0.034744525547445254,
+      "grad_norm": 1.830377999961128,
+      "learning_rate": 9.999427636175228e-06,
+      "loss": 0.4297,
+      "step": 357
+    },
+    {
+      "epoch": 0.034841849148418494,
+      "grad_norm": 2.1217123292302875,
+      "learning_rate": 9.999403539739837e-06,
+      "loss": 0.3605,
+      "step": 358
+    },
+    {
+      "epoch": 0.03493917274939173,
+      "grad_norm": 2.001599625802253,
+      "learning_rate": 9.999378946510642e-06,
+      "loss": 0.5237,
+      "step": 359
+    },
+    {
+      "epoch": 0.035036496350364967,
+      "grad_norm": 1.6719956399048532,
+      "learning_rate": 9.99935385649009e-06,
+      "loss": 0.424,
+      "step": 360
+    },
+    {
+      "epoch": 0.0351338199513382,
+      "grad_norm": 1.5962062682133515,
+      "learning_rate": 9.99932826968067e-06,
+      "loss": 0.4228,
+      "step": 361
+    },
+    {
+      "epoch": 0.03523114355231143,
+      "grad_norm": 1.9754274750693919,
+      "learning_rate": 9.999302186084929e-06,
+      "loss": 0.4333,
+      "step": 362
+    },
+    {
+      "epoch": 0.03532846715328467,
+      "grad_norm": 1.8248617929879183,
+      "learning_rate": 9.999275605705457e-06,
+      "loss": 0.4985,
+      "step": 363
+    },
+    {
+      "epoch": 0.035425790754257905,
+      "grad_norm": 2.5923075514224982,
+      "learning_rate": 9.999248528544895e-06,
+      "loss": 0.4829,
+      "step": 364
+    },
+    {
+      "epoch": 0.035523114355231145,
+      "grad_norm": 1.9900801938638135,
+      "learning_rate": 9.999220954605932e-06,
+      "loss": 0.587,
+      "step": 365
+    },
+    {
+      "epoch": 0.03562043795620438,
+      "grad_norm": 1.731558772897005,
+      "learning_rate": 9.999192883891314e-06,
+      "loss": 0.3299,
+      "step": 366
+    },
+    {
+      "epoch": 0.03571776155717762,
+      "grad_norm": 2.339577788711278,
+      "learning_rate": 9.999164316403823e-06,
+      "loss": 0.4845,
+      "step": 367
+    },
+    {
+      "epoch": 0.03581508515815085,
+      "grad_norm": 1.9784113864985955,
+      "learning_rate": 9.999135252146302e-06,
+      "loss": 0.5776,
+      "step": 368
+    },
+    {
+      "epoch": 0.03591240875912409,
+      "grad_norm": 1.5555461256937277,
+      "learning_rate": 9.999105691121638e-06,
+      "loss": 0.3563,
+      "step": 369
+    },
+    {
+      "epoch": 0.03600973236009732,
+      "grad_norm": 1.7905677559908044,
+      "learning_rate": 9.99907563333277e-06,
+      "loss": 0.546,
+      "step": 370
+    },
+    {
+      "epoch": 0.03610705596107056,
+      "grad_norm": 2.0490894714600287,
+      "learning_rate": 9.999045078782684e-06,
+      "loss": 0.6836,
+      "step": 371
+    },
+    {
+      "epoch": 0.036204379562043795,
+      "grad_norm": 2.216601446334751,
+      "learning_rate": 9.999014027474413e-06,
+      "loss": 0.5237,
+      "step": 372
+    },
+    {
+      "epoch": 0.036301703163017035,
+      "grad_norm": 1.5937926342815392,
+      "learning_rate": 9.998982479411047e-06,
+      "loss": 0.3539,
+      "step": 373
+    },
+    {
+      "epoch": 0.03639902676399027,
+      "grad_norm": 2.3941848280266864,
+      "learning_rate": 9.99895043459572e-06,
+      "loss": 0.6249,
+      "step": 374
+    },
+    {
+      "epoch": 0.0364963503649635,
+      "grad_norm": 2.072859669066288,
+      "learning_rate": 9.998917893031615e-06,
+      "loss": 0.5415,
+      "step": 375
+    },
+    {
+      "epoch": 0.03659367396593674,
+      "grad_norm": 1.670908711065728,
+      "learning_rate": 9.998884854721968e-06,
+      "loss": 0.3034,
+      "step": 376
+    },
+    {
+      "epoch": 0.03669099756690997,
+      "grad_norm": 1.9880303784818283,
+      "learning_rate": 9.998851319670057e-06,
+      "loss": 0.5025,
+      "step": 377
+    },
+    {
+      "epoch": 0.03678832116788321,
+      "grad_norm": 1.3517666701087396,
+      "learning_rate": 9.99881728787922e-06,
+      "loss": 0.2775,
+      "step": 378
+    },
+    {
+      "epoch": 0.036885644768856446,
+      "grad_norm": 1.8952553535268069,
+      "learning_rate": 9.998782759352839e-06,
+      "loss": 0.5306,
+      "step": 379
+    },
+    {
+      "epoch": 0.036982968369829686,
+      "grad_norm": 1.8730537486024816,
+      "learning_rate": 9.998747734094338e-06,
+      "loss": 0.386,
+      "step": 380
+    },
+    {
+      "epoch": 0.03708029197080292,
+      "grad_norm": 2.058996056292158,
+      "learning_rate": 9.998712212107205e-06,
+      "loss": 0.5641,
+      "step": 381
+    },
+    {
+      "epoch": 0.03717761557177616,
+      "grad_norm": 1.9837834234853275,
+      "learning_rate": 9.998676193394966e-06,
+      "loss": 0.2628,
+      "step": 382
+    },
+    {
+      "epoch": 0.03727493917274939,
+      "grad_norm": 2.189700953999047,
+      "learning_rate": 9.998639677961203e-06,
+      "loss": 0.6024,
+      "step": 383
+    },
+    {
+      "epoch": 0.03737226277372263,
+      "grad_norm": 2.060696593716547,
+      "learning_rate": 9.99860266580954e-06,
+      "loss": 0.5377,
+      "step": 384
+    },
+    {
+      "epoch": 0.037469586374695864,
+      "grad_norm": 2.0831966609629227,
+      "learning_rate": 9.99856515694366e-06,
+      "loss": 0.5063,
+      "step": 385
+    },
+    {
+      "epoch": 0.0375669099756691,
+      "grad_norm": 2.2950496556846227,
+      "learning_rate": 9.998527151367288e-06,
+      "loss": 0.6484,
+      "step": 386
+    },
+    {
+      "epoch": 0.037664233576642336,
+      "grad_norm": 2.2597922123273873,
+      "learning_rate": 9.9984886490842e-06,
+      "loss": 0.6617,
+      "step": 387
+    },
+    {
+      "epoch": 0.03776155717761557,
+      "grad_norm": 2.071575887731456,
+      "learning_rate": 9.99844965009822e-06,
+      "loss": 0.5405,
+      "step": 388
+    },
+    {
+      "epoch": 0.03785888077858881,
+      "grad_norm": 2.004249587957457,
+      "learning_rate": 9.99841015441323e-06,
+      "loss": 0.4306,
+      "step": 389
+    },
+    {
+      "epoch": 0.03795620437956204,
+      "grad_norm": 1.9297023880727862,
+      "learning_rate": 9.99837016203315e-06,
+      "loss": 0.4083,
+      "step": 390
+    },
+    {
+      "epoch": 0.03805352798053528,
+      "grad_norm": 2.001337081282171,
+      "learning_rate": 9.998329672961952e-06,
+      "loss": 0.4999,
+      "step": 391
+    },
+    {
+      "epoch": 0.038150851581508514,
+      "grad_norm": 1.7630230797021285,
+      "learning_rate": 9.998288687203665e-06,
+      "loss": 0.4267,
+      "step": 392
+    },
+    {
+      "epoch": 0.038248175182481754,
+      "grad_norm": 1.4413546421147376,
+      "learning_rate": 9.998247204762358e-06,
+      "loss": 0.3028,
+      "step": 393
+    },
+    {
+      "epoch": 0.03834549878345499,
+      "grad_norm": 2.032450629241147,
+      "learning_rate": 9.998205225642154e-06,
+      "loss": 0.4216,
+      "step": 394
+    },
+    {
+      "epoch": 0.03844282238442823,
+      "grad_norm": 1.8288270303352272,
+      "learning_rate": 9.998162749847224e-06,
+      "loss": 0.451,
+      "step": 395
+    },
+    {
+      "epoch": 0.03854014598540146,
+      "grad_norm": 1.5869427581540143,
+      "learning_rate": 9.998119777381791e-06,
+      "loss": 0.4896,
+      "step": 396
+    },
+    {
+      "epoch": 0.03863746958637469,
+      "grad_norm": 1.9312614168983935,
+      "learning_rate": 9.998076308250122e-06,
+      "loss": 0.351,
+      "step": 397
+    },
+    {
+      "epoch": 0.03873479318734793,
+      "grad_norm": 2.182734939846557,
+      "learning_rate": 9.99803234245654e-06,
+      "loss": 0.4456,
+      "step": 398
+    },
+    {
+      "epoch": 0.038832116788321165,
+      "grad_norm": 1.6075130172605856,
+      "learning_rate": 9.997987880005412e-06,
+      "loss": 0.3333,
+      "step": 399
+    },
+    {
+      "epoch": 0.038929440389294405,
+      "grad_norm": 2.0206579020801048,
+      "learning_rate": 9.997942920901154e-06,
+      "loss": 0.4662,
+      "step": 400
+    },
+    {
+      "epoch": 0.03902676399026764,
+      "grad_norm": 2.0019154912621246,
+      "learning_rate": 9.997897465148236e-06,
+      "loss": 0.588,
+      "step": 401
+    },
+    {
+      "epoch": 0.03912408759124088,
+      "grad_norm": 1.9556688755730123,
+      "learning_rate": 9.997851512751178e-06,
+      "loss": 0.5364,
+      "step": 402
+    },
+    {
+      "epoch": 0.03922141119221411,
+      "grad_norm": 2.1735940620422687,
+      "learning_rate": 9.997805063714541e-06,
+      "loss": 0.4155,
+      "step": 403
+    },
+    {
+      "epoch": 0.03931873479318735,
+      "grad_norm": 1.893104755523836,
+      "learning_rate": 9.997758118042945e-06,
+      "loss": 0.2835,
+      "step": 404
+    },
+    {
+      "epoch": 0.03941605839416058,
+      "grad_norm": 1.892857392200546,
+      "learning_rate": 9.99771067574105e-06,
+      "loss": 0.317,
+      "step": 405
+    },
+    {
+      "epoch": 0.03951338199513382,
+      "grad_norm": 2.194365925195629,
+      "learning_rate": 9.997662736813575e-06,
+      "loss": 0.5972,
+      "step": 406
+    },
+    {
+      "epoch": 0.039610705596107056,
+      "grad_norm": 2.3359516870584547,
+      "learning_rate": 9.997614301265281e-06,
+      "loss": 0.3505,
+      "step": 407
+    },
+    {
+      "epoch": 0.039708029197080295,
+      "grad_norm": 1.8041349283411827,
+      "learning_rate": 9.997565369100983e-06,
+      "loss": 0.4003,
+      "step": 408
+    },
+    {
+      "epoch": 0.03980535279805353,
+      "grad_norm": 2.2199870140108273,
+      "learning_rate": 9.997515940325542e-06,
+      "loss": 0.4428,
+      "step": 409
+    },
+    {
+      "epoch": 0.03990267639902676,
+      "grad_norm": 2.193796849633566,
+      "learning_rate": 9.997466014943871e-06,
+      "loss": 0.3906,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7309920828616168,
+      "learning_rate": 9.99741559296093e-06,
+      "loss": 0.6283,
+      "step": 411
+    },
+    {
+      "epoch": 0.040097323600973234,
+      "grad_norm": 2.220745639846989,
+      "learning_rate": 9.99736467438173e-06,
+      "loss": 0.4568,
+      "step": 412
+    },
+    {
+      "epoch": 0.04019464720194647,
+      "grad_norm": 1.905067765139487,
+      "learning_rate": 9.99731325921133e-06,
+      "loss": 0.3198,
+      "step": 413
+    },
+    {
+      "epoch": 0.040291970802919706,
+      "grad_norm": 2.0461180940034116,
+      "learning_rate": 9.997261347454841e-06,
+      "loss": 0.3783,
+      "step": 414
+    },
+    {
+      "epoch": 0.040389294403892946,
+      "grad_norm": 1.9732614929529544,
+      "learning_rate": 9.99720893911742e-06,
+      "loss": 0.5211,
+      "step": 415
+    },
+    {
+      "epoch": 0.04048661800486618,
+      "grad_norm": 2.341156401873798,
+      "learning_rate": 9.997156034204276e-06,
+      "loss": 0.5094,
+      "step": 416
+    },
+    {
+      "epoch": 0.04058394160583942,
+      "grad_norm": 2.2588135503158138,
+      "learning_rate": 9.997102632720664e-06,
+      "loss": 0.591,
+      "step": 417
+    },
+    {
+      "epoch": 0.04068126520681265,
+      "grad_norm": 2.187795564574772,
+      "learning_rate": 9.997048734671893e-06,
+      "loss": 0.3811,
+      "step": 418
+    },
+    {
+      "epoch": 0.04077858880778589,
+      "grad_norm": 2.2570398189900938,
+      "learning_rate": 9.996994340063314e-06,
+      "loss": 0.4494,
+      "step": 419
+    },
+    {
+      "epoch": 0.040875912408759124,
+      "grad_norm": 2.3267878846596597,
+      "learning_rate": 9.996939448900341e-06,
+      "loss": 0.5254,
+      "step": 420
+    },
+    {
+      "epoch": 0.04097323600973236,
+      "grad_norm": 1.9149387144635641,
+      "learning_rate": 9.99688406118842e-06,
+      "loss": 0.4281,
+      "step": 421
+    },
+    {
+      "epoch": 0.0410705596107056,
+      "grad_norm": 2.4052095021382285,
+      "learning_rate": 9.996828176933062e-06,
+      "loss": 0.61,
+      "step": 422
+    },
+    {
+      "epoch": 0.04116788321167883,
+      "grad_norm": 2.8744864627123237,
+      "learning_rate": 9.996771796139814e-06,
+      "loss": 0.4708,
+      "step": 423
+    },
+    {
+      "epoch": 0.04126520681265207,
+      "grad_norm": 2.0334953222734513,
+      "learning_rate": 9.996714918814284e-06,
+      "loss": 0.2697,
+      "step": 424
+    },
+    {
+      "epoch": 0.0413625304136253,
+      "grad_norm": 2.1314093477075486,
+      "learning_rate": 9.996657544962119e-06,
+      "loss": 0.3026,
+      "step": 425
+    },
+    {
+      "epoch": 0.04145985401459854,
+      "grad_norm": 1.7241742631767316,
+      "learning_rate": 9.996599674589022e-06,
+      "loss": 0.3624,
+      "step": 426
+    },
+    {
+      "epoch": 0.041557177615571775,
+      "grad_norm": 2.417754377928955,
+      "learning_rate": 9.996541307700746e-06,
+      "loss": 0.6682,
+      "step": 427
+    },
+    {
+      "epoch": 0.041654501216545015,
+      "grad_norm": 2.2126055245100256,
+      "learning_rate": 9.99648244430309e-06,
+      "loss": 0.3705,
+      "step": 428
+    },
+    {
+      "epoch": 0.04175182481751825,
+      "grad_norm": 1.8224510106748588,
+      "learning_rate": 9.996423084401901e-06,
+      "loss": 0.4318,
+      "step": 429
+    },
+    {
+      "epoch": 0.04184914841849149,
+      "grad_norm": 1.6786428352287364,
+      "learning_rate": 9.996363228003079e-06,
+      "loss": 0.4662,
+      "step": 430
+    },
+    {
+      "epoch": 0.04194647201946472,
+      "grad_norm": 1.9342922605897592,
+      "learning_rate": 9.99630287511257e-06,
+      "loss": 0.4874,
+      "step": 431
+    },
+    {
+      "epoch": 0.04204379562043795,
+      "grad_norm": 1.9444011100602645,
+      "learning_rate": 9.996242025736377e-06,
+      "loss": 0.3711,
+      "step": 432
+    },
+    {
+      "epoch": 0.04214111922141119,
+      "grad_norm": 3.114184163688958,
+      "learning_rate": 9.99618067988054e-06,
+      "loss": 0.5342,
+      "step": 433
+    },
+    {
+      "epoch": 0.042238442822384425,
+      "grad_norm": 1.993932460938173,
+      "learning_rate": 9.99611883755116e-06,
+      "loss": 0.465,
+      "step": 434
+    },
+    {
+      "epoch": 0.042335766423357665,
+      "grad_norm": 1.5062408953506277,
+      "learning_rate": 9.99605649875438e-06,
+      "loss": 0.3862,
+      "step": 435
+    },
+    {
+      "epoch": 0.0424330900243309,
+      "grad_norm": 2.5287447175721733,
+      "learning_rate": 9.995993663496394e-06,
+      "loss": 0.5638,
+      "step": 436
+    },
+    {
+      "epoch": 0.04253041362530414,
+      "grad_norm": 1.7215400937807486,
+      "learning_rate": 9.995930331783448e-06,
+      "loss": 0.3507,
+      "step": 437
+    },
+    {
+      "epoch": 0.04262773722627737,
+      "grad_norm": 1.5105936757865817,
+      "learning_rate": 9.995866503621834e-06,
+      "loss": 0.4086,
+      "step": 438
+    },
+    {
+      "epoch": 0.04272506082725061,
+      "grad_norm": 1.828501540310894,
+      "learning_rate": 9.995802179017893e-06,
+      "loss": 0.3477,
+      "step": 439
+    },
+    {
+      "epoch": 0.04282238442822384,
+      "grad_norm": 1.6658361590948114,
+      "learning_rate": 9.995737357978022e-06,
+      "loss": 0.4006,
+      "step": 440
+    },
+    {
+      "epoch": 0.04291970802919708,
+      "grad_norm": 1.6434395036324305,
+      "learning_rate": 9.995672040508656e-06,
+      "loss": 0.4349,
+      "step": 441
+    },
+    {
+      "epoch": 0.043017031630170316,
+      "grad_norm": 1.9913424027071125,
+      "learning_rate": 9.99560622661629e-06,
+      "loss": 0.3415,
+      "step": 442
+    },
+    {
+      "epoch": 0.043114355231143556,
+      "grad_norm": 1.6487474195389296,
+      "learning_rate": 9.995539916307463e-06,
+      "loss": 0.4804,
+      "step": 443
+    },
+    {
+      "epoch": 0.04321167883211679,
+      "grad_norm": 1.4861266391850032,
+      "learning_rate": 9.995473109588764e-06,
+      "loss": 0.411,
+      "step": 444
+    },
+    {
+      "epoch": 0.04330900243309002,
+      "grad_norm": 1.4390762643228305,
+      "learning_rate": 9.995405806466831e-06,
+      "loss": 0.3806,
+      "step": 445
+    },
+    {
+      "epoch": 0.04340632603406326,
+      "grad_norm": 1.7775332171720517,
+      "learning_rate": 9.995338006948353e-06,
+      "loss": 0.3332,
+      "step": 446
+    },
+    {
+      "epoch": 0.043503649635036494,
+      "grad_norm": 1.7312883283317864,
+      "learning_rate": 9.995269711040067e-06,
+      "loss": 0.2736,
+      "step": 447
+    },
+    {
+      "epoch": 0.043600973236009734,
+      "grad_norm": 1.7973901424872405,
+      "learning_rate": 9.995200918748759e-06,
+      "loss": 0.5597,
+      "step": 448
+    },
+    {
+      "epoch": 0.04369829683698297,
+      "grad_norm": 2.0409413301370334,
+      "learning_rate": 9.995131630081265e-06,
+      "loss": 0.6045,
+      "step": 449
+    },
+    {
+      "epoch": 0.043795620437956206,
+      "grad_norm": 3.2708903670147347,
+      "learning_rate": 9.995061845044473e-06,
+      "loss": 0.6245,
+      "step": 450
+    },
+    {
+      "epoch": 0.04389294403892944,
+      "grad_norm": 1.744466889932859,
+      "learning_rate": 9.994991563645314e-06,
+      "loss": 0.4129,
+      "step": 451
+    },
+    {
+      "epoch": 0.04399026763990268,
+      "grad_norm": 1.8775864246251477,
+      "learning_rate": 9.994920785890771e-06,
+      "loss": 0.414,
+      "step": 452
+    },
+    {
+      "epoch": 0.04408759124087591,
+      "grad_norm": 1.3868286948878126,
+      "learning_rate": 9.994849511787881e-06,
+      "loss": 0.3164,
+      "step": 453
+    },
+    {
+      "epoch": 0.04418491484184915,
+      "grad_norm": 1.6888257223301795,
+      "learning_rate": 9.994777741343727e-06,
+      "loss": 0.3241,
+      "step": 454
+    },
+    {
+      "epoch": 0.044282238442822384,
+      "grad_norm": 1.5029594314338663,
+      "learning_rate": 9.994705474565436e-06,
+      "loss": 0.4148,
+      "step": 455
+    },
+    {
+      "epoch": 0.04437956204379562,
+      "grad_norm": 1.7159996915963702,
+      "learning_rate": 9.994632711460193e-06,
+      "loss": 0.3387,
+      "step": 456
+    },
+    {
+      "epoch": 0.04447688564476886,
+      "grad_norm": 1.7717997513120352,
+      "learning_rate": 9.994559452035228e-06,
+      "loss": 0.4547,
+      "step": 457
+    },
+    {
+      "epoch": 0.04457420924574209,
+      "grad_norm": 1.887765282184233,
+      "learning_rate": 9.99448569629782e-06,
+      "loss": 0.5919,
+      "step": 458
+    },
+    {
+      "epoch": 0.04467153284671533,
+      "grad_norm": 2.0151049512314585,
+      "learning_rate": 9.994411444255298e-06,
+      "loss": 0.4556,
+      "step": 459
+    },
+    {
+      "epoch": 0.04476885644768856,
+      "grad_norm": 1.5706463359289826,
+      "learning_rate": 9.994336695915041e-06,
+      "loss": 0.3443,
+      "step": 460
+    },
+    {
+      "epoch": 0.0448661800486618,
+      "grad_norm": 1.9067884841542395,
+      "learning_rate": 9.994261451284477e-06,
+      "loss": 0.5862,
+      "step": 461
+    },
+    {
+      "epoch": 0.044963503649635035,
+      "grad_norm": 1.7346846845298518,
+      "learning_rate": 9.994185710371083e-06,
+      "loss": 0.3588,
+      "step": 462
+    },
+    {
+      "epoch": 0.045060827250608275,
+      "grad_norm": 1.5593715629463312,
+      "learning_rate": 9.994109473182385e-06,
+      "loss": 0.2891,
+      "step": 463
+    },
+    {
+      "epoch": 0.04515815085158151,
+      "grad_norm": 2.326736753149576,
+      "learning_rate": 9.994032739725959e-06,
+      "loss": 0.6517,
+      "step": 464
+    },
+    {
+      "epoch": 0.04525547445255475,
+      "grad_norm": 2.2142852132770305,
+      "learning_rate": 9.99395551000943e-06,
+      "loss": 0.3571,
+      "step": 465
+    },
+    {
+      "epoch": 0.04535279805352798,
+      "grad_norm": 1.7351954813390544,
+      "learning_rate": 9.993877784040474e-06,
+      "loss": 0.3849,
+      "step": 466
+    },
+    {
+      "epoch": 0.04545012165450121,
+      "grad_norm": 1.3962336815381617,
+      "learning_rate": 9.993799561826811e-06,
+      "loss": 0.311,
+      "step": 467
+    },
+    {
+      "epoch": 0.04554744525547445,
+      "grad_norm": 1.878958465421645,
+      "learning_rate": 9.993720843376216e-06,
+      "loss": 0.5602,
+      "step": 468
+    },
+    {
+      "epoch": 0.045644768856447686,
+      "grad_norm": 1.519160992933857,
+      "learning_rate": 9.993641628696513e-06,
+      "loss": 0.2379,
+      "step": 469
+    },
+    {
+      "epoch": 0.045742092457420926,
+      "grad_norm": 2.5345930464298885,
+      "learning_rate": 9.99356191779557e-06,
+      "loss": 0.4239,
+      "step": 470
+    },
+    {
+      "epoch": 0.04583941605839416,
+      "grad_norm": 1.3153911718041251,
+      "learning_rate": 9.993481710681314e-06,
+      "loss": 0.3454,
+      "step": 471
+    },
+    {
+      "epoch": 0.0459367396593674,
+      "grad_norm": 2.16208125563947,
+      "learning_rate": 9.993401007361707e-06,
+      "loss": 0.5386,
+      "step": 472
+    },
+    {
+      "epoch": 0.04603406326034063,
+      "grad_norm": 1.8150842593472827,
+      "learning_rate": 9.993319807844775e-06,
+      "loss": 0.3077,
+      "step": 473
+    },
+    {
+      "epoch": 0.04613138686131387,
+      "grad_norm": 1.6656864462678063,
+      "learning_rate": 9.993238112138584e-06,
+      "loss": 0.4927,
+      "step": 474
+    },
+    {
+      "epoch": 0.046228710462287104,
+      "grad_norm": 1.3429917702468868,
+      "learning_rate": 9.993155920251252e-06,
+      "loss": 0.2433,
+      "step": 475
+    },
+    {
+      "epoch": 0.04632603406326034,
+      "grad_norm": 1.3651155739367906,
+      "learning_rate": 9.993073232190949e-06,
+      "loss": 0.2947,
+      "step": 476
+    },
+    {
+      "epoch": 0.046423357664233576,
+      "grad_norm": 1.7815516701613203,
+      "learning_rate": 9.992990047965887e-06,
+      "loss": 0.5372,
+      "step": 477
+    },
+    {
+      "epoch": 0.046520681265206816,
+      "grad_norm": 1.846696342179327,
+      "learning_rate": 9.992906367584337e-06,
+      "loss": 0.5127,
+      "step": 478
+    },
+    {
+      "epoch": 0.04661800486618005,
+      "grad_norm": 1.7511253825578088,
+      "learning_rate": 9.992822191054612e-06,
+      "loss": 0.4074,
+      "step": 479
+    },
+    {
+      "epoch": 0.04671532846715328,
+      "grad_norm": 1.8105635986872588,
+      "learning_rate": 9.992737518385076e-06,
+      "loss": 0.4998,
+      "step": 480
+    },
+    {
+      "epoch": 0.04681265206812652,
+      "grad_norm": 2.2743597617900746,
+      "learning_rate": 9.992652349584147e-06,
+      "loss": 0.6249,
+      "step": 481
+    },
+    {
+      "epoch": 0.046909975669099754,
+      "grad_norm": 1.93948496382319,
+      "learning_rate": 9.992566684660282e-06,
+      "loss": 0.5411,
+      "step": 482
+    },
+    {
+      "epoch": 0.047007299270072994,
+      "grad_norm": 1.4073760716303516,
+      "learning_rate": 9.992480523621999e-06,
+      "loss": 0.3506,
+      "step": 483
+    },
+    {
+      "epoch": 0.04710462287104623,
+      "grad_norm": 1.388293079160528,
+      "learning_rate": 9.992393866477856e-06,
+      "loss": 0.3304,
+      "step": 484
+    },
+    {
+      "epoch": 0.04720194647201947,
+      "grad_norm": 2.082643572745618,
+      "learning_rate": 9.992306713236467e-06,
+      "loss": 0.5653,
+      "step": 485
+    },
+    {
+      "epoch": 0.0472992700729927,
+      "grad_norm": 1.7104664332606834,
+      "learning_rate": 9.992219063906492e-06,
+      "loss": 0.3317,
+      "step": 486
+    },
+    {
+      "epoch": 0.04739659367396594,
+      "grad_norm": 1.7575848919482624,
+      "learning_rate": 9.992130918496638e-06,
+      "loss": 0.4109,
+      "step": 487
+    },
+    {
+      "epoch": 0.04749391727493917,
+      "grad_norm": 1.7351379091271637,
+      "learning_rate": 9.992042277015668e-06,
+      "loss": 0.5065,
+      "step": 488
+    },
+    {
+      "epoch": 0.04759124087591241,
+      "grad_norm": 1.4444570948381004,
+      "learning_rate": 9.991953139472387e-06,
+      "loss": 0.4023,
+      "step": 489
+    },
+    {
+      "epoch": 0.047688564476885645,
+      "grad_norm": 1.4697709289140384,
+      "learning_rate": 9.991863505875656e-06,
+      "loss": 0.3364,
+      "step": 490
+    },
+    {
+      "epoch": 0.04778588807785888,
+      "grad_norm": 1.9428205960506804,
+      "learning_rate": 9.99177337623438e-06,
+      "loss": 0.4303,
+      "step": 491
+    },
+    {
+      "epoch": 0.04788321167883212,
+      "grad_norm": 1.931152158561148,
+      "learning_rate": 9.991682750557516e-06,
+      "loss": 0.2857,
+      "step": 492
+    },
+    {
+      "epoch": 0.04798053527980535,
+      "grad_norm": 1.9301394655308035,
+      "learning_rate": 9.991591628854067e-06,
+      "loss": 0.5998,
+      "step": 493
+    },
+    {
+      "epoch": 0.04807785888077859,
+      "grad_norm": 1.7788293016868693,
+      "learning_rate": 9.99150001113309e-06,
+      "loss": 0.4595,
+      "step": 494
+    },
+    {
+      "epoch": 0.04817518248175182,
+      "grad_norm": 2.0641225732440134,
+      "learning_rate": 9.99140789740369e-06,
+      "loss": 0.3848,
+      "step": 495
+    },
+    {
+      "epoch": 0.04827250608272506,
+      "grad_norm": 2.2832955373527044,
+      "learning_rate": 9.99131528767502e-06,
+      "loss": 0.6396,
+      "step": 496
+    },
+    {
+      "epoch": 0.048369829683698295,
+      "grad_norm": 1.6658790952812916,
+      "learning_rate": 9.99122218195628e-06,
+      "loss": 0.5429,
+      "step": 497
+    },
+    {
+      "epoch": 0.048467153284671535,
+      "grad_norm": 1.6568038302360257,
+      "learning_rate": 9.991128580256725e-06,
+      "loss": 0.4532,
+      "step": 498
+    },
+    {
+      "epoch": 0.04856447688564477,
+      "grad_norm": 1.8451659374514144,
+      "learning_rate": 9.991034482585656e-06,
+      "loss": 0.5845,
+      "step": 499
+    },
+    {
+      "epoch": 0.04866180048661801,
+      "grad_norm": 1.9103948838029656,
+      "learning_rate": 9.99093988895242e-06,
+      "loss": 0.5508,
+      "step": 500
+    },
+    {
+      "epoch": 0.04875912408759124,
+      "grad_norm": 1.9691733858712537,
+      "learning_rate": 9.990844799366422e-06,
+      "loss": 0.6374,
+      "step": 501
+    },
+    {
+      "epoch": 0.048856447688564474,
+      "grad_norm": 2.1278472226161846,
+      "learning_rate": 9.990749213837108e-06,
+      "loss": 0.572,
+      "step": 502
+    },
+    {
+      "epoch": 0.04895377128953771,
+      "grad_norm": 1.9704028865885994,
+      "learning_rate": 9.990653132373977e-06,
+      "loss": 0.6282,
+      "step": 503
+    },
+    {
+      "epoch": 0.049051094890510946,
+      "grad_norm": 1.8965741341561362,
+      "learning_rate": 9.990556554986577e-06,
+      "loss": 0.5749,
+      "step": 504
+    },
+    {
+      "epoch": 0.049148418491484186,
+      "grad_norm": 1.5425018763105707,
+      "learning_rate": 9.990459481684504e-06,
+      "loss": 0.4236,
+      "step": 505
+    },
+    {
+      "epoch": 0.04924574209245742,
+      "grad_norm": 1.736669998068125,
+      "learning_rate": 9.990361912477405e-06,
+      "loss": 0.4275,
+      "step": 506
+    },
+    {
+      "epoch": 0.04934306569343066,
+      "grad_norm": 2.049335776858506,
+      "learning_rate": 9.990263847374976e-06,
+      "loss": 0.6897,
+      "step": 507
+    },
+    {
+      "epoch": 0.04944038929440389,
+      "grad_norm": 1.8544975871268152,
+      "learning_rate": 9.990165286386961e-06,
+      "loss": 0.4811,
+      "step": 508
+    },
+    {
+      "epoch": 0.04953771289537713,
+      "grad_norm": 1.5709178763522822,
+      "learning_rate": 9.990066229523155e-06,
+      "loss": 0.4585,
+      "step": 509
+    },
+    {
+      "epoch": 0.049635036496350364,
+      "grad_norm": 2.1410068811754153,
+      "learning_rate": 9.989966676793399e-06,
+      "loss": 0.4773,
+      "step": 510
+    },
+    {
+      "epoch": 0.049732360097323604,
+      "grad_norm": 1.760724042734433,
+      "learning_rate": 9.989866628207589e-06,
+      "loss": 0.3144,
+      "step": 511
+    },
+    {
+      "epoch": 0.04982968369829684,
+      "grad_norm": 1.8521560168370175,
+      "learning_rate": 9.989766083775662e-06,
+      "loss": 0.4656,
+      "step": 512
+    },
+    {
+      "epoch": 0.049927007299270076,
+      "grad_norm": 1.544987615640627,
+      "learning_rate": 9.989665043507616e-06,
+      "loss": 0.4089,
+      "step": 513
+    },
+    {
+      "epoch": 0.05002433090024331,
+      "grad_norm": 1.9122960249889975,
+      "learning_rate": 9.989563507413487e-06,
+      "loss": 0.4535,
+      "step": 514
+    },
+    {
+      "epoch": 0.05012165450121654,
+      "grad_norm": 1.5187134098621655,
+      "learning_rate": 9.989461475503363e-06,
+      "loss": 0.31,
+      "step": 515
+    },
+    {
+      "epoch": 0.05021897810218978,
+      "grad_norm": 1.562160455050312,
+      "learning_rate": 9.989358947787389e-06,
+      "loss": 0.4009,
+      "step": 516
+    },
+    {
+      "epoch": 0.050316301703163015,
+      "grad_norm": 1.738084966314413,
+      "learning_rate": 9.989255924275746e-06,
+      "loss": 0.4723,
+      "step": 517
+    },
+    {
+      "epoch": 0.050413625304136254,
+      "grad_norm": 2.156580581755068,
+      "learning_rate": 9.989152404978678e-06,
+      "loss": 0.4407,
+      "step": 518
+    },
+    {
+      "epoch": 0.05051094890510949,
+      "grad_norm": 1.8652302207700793,
+      "learning_rate": 9.989048389906469e-06,
+      "loss": 0.587,
+      "step": 519
+    },
+    {
+      "epoch": 0.05060827250608273,
+      "grad_norm": 1.5934369396830426,
+      "learning_rate": 9.988943879069452e-06,
+      "loss": 0.3961,
+      "step": 520
+    },
+    {
+      "epoch": 0.05070559610705596,
+      "grad_norm": 1.4294562647861604,
+      "learning_rate": 9.988838872478017e-06,
+      "loss": 0.3382,
+      "step": 521
+    },
+    {
+      "epoch": 0.0508029197080292,
+      "grad_norm": 1.5693240874435923,
+      "learning_rate": 9.988733370142598e-06,
+      "loss": 0.3876,
+      "step": 522
+    },
+    {
+      "epoch": 0.05090024330900243,
+      "grad_norm": 1.6720738515514542,
+      "learning_rate": 9.988627372073678e-06,
+      "loss": 0.448,
+      "step": 523
+    },
+    {
+      "epoch": 0.05099756690997567,
+      "grad_norm": 2.0438207304961367,
+      "learning_rate": 9.988520878281787e-06,
+      "loss": 0.5724,
+      "step": 524
+    },
+    {
+      "epoch": 0.051094890510948905,
+      "grad_norm": 2.0003463921985456,
+      "learning_rate": 9.988413888777512e-06,
+      "loss": 0.4506,
+      "step": 525
+    },
+    {
+      "epoch": 0.05119221411192214,
+      "grad_norm": 2.11812759304704,
+      "learning_rate": 9.988306403571482e-06,
+      "loss": 0.757,
+      "step": 526
+    },
+    {
+      "epoch": 0.05128953771289538,
+      "grad_norm": 1.5594386055307068,
+      "learning_rate": 9.98819842267438e-06,
+      "loss": 0.4145,
+      "step": 527
+    },
+    {
+      "epoch": 0.05138686131386861,
+      "grad_norm": 1.917978943216931,
+      "learning_rate": 9.988089946096933e-06,
+      "loss": 0.5363,
+      "step": 528
+    },
+    {
+      "epoch": 0.05148418491484185,
+      "grad_norm": 1.3212282063862113,
+      "learning_rate": 9.987980973849924e-06,
+      "loss": 0.3132,
+      "step": 529
+    },
+    {
+      "epoch": 0.05158150851581508,
+      "grad_norm": 1.2285769982465171,
+      "learning_rate": 9.987871505944177e-06,
+      "loss": 0.2287,
+      "step": 530
+    },
+    {
+      "epoch": 0.05167883211678832,
+      "grad_norm": 1.849610792922833,
+      "learning_rate": 9.987761542390574e-06,
+      "loss": 0.6487,
+      "step": 531
+    },
+    {
+      "epoch": 0.051776155717761556,
+      "grad_norm": 1.158461389164102,
+      "learning_rate": 9.987651083200044e-06,
+      "loss": 0.2111,
+      "step": 532
+    },
+    {
+      "epoch": 0.051873479318734796,
+      "grad_norm": 1.8450520976911682,
+      "learning_rate": 9.987540128383556e-06,
+      "loss": 0.5579,
+      "step": 533
+    },
+    {
+      "epoch": 0.05197080291970803,
+      "grad_norm": 1.9047794610871986,
+      "learning_rate": 9.98742867795214e-06,
+      "loss": 0.4542,
+      "step": 534
+    },
+    {
+      "epoch": 0.05206812652068127,
+      "grad_norm": 1.5564676952152843,
+      "learning_rate": 9.987316731916872e-06,
+      "loss": 0.4467,
+      "step": 535
+    },
+    {
+      "epoch": 0.0521654501216545,
+      "grad_norm": 1.403952395827601,
+      "learning_rate": 9.987204290288876e-06,
+      "loss": 0.3761,
+      "step": 536
+    },
+    {
+      "epoch": 0.052262773722627734,
+      "grad_norm": 1.948151749349848,
+      "learning_rate": 9.987091353079323e-06,
+      "loss": 0.5782,
+      "step": 537
+    },
+    {
+      "epoch": 0.052360097323600974,
+      "grad_norm": 1.6211222818460531,
+      "learning_rate": 9.986977920299437e-06,
+      "loss": 0.4047,
+      "step": 538
+    },
+    {
+      "epoch": 0.052457420924574207,
+      "grad_norm": 1.4911900726837217,
+      "learning_rate": 9.986863991960491e-06,
+      "loss": 0.3817,
+      "step": 539
+    },
+    {
+      "epoch": 0.052554744525547446,
+      "grad_norm": 1.530872687739145,
+      "learning_rate": 9.986749568073804e-06,
+      "loss": 0.4639,
+      "step": 540
+    },
+    {
+      "epoch": 0.05265206812652068,
+      "grad_norm": 1.766399180057757,
+      "learning_rate": 9.986634648650746e-06,
+      "loss": 0.5132,
+      "step": 541
+    },
+    {
+      "epoch": 0.05274939172749392,
+      "grad_norm": 1.7318370911583716,
+      "learning_rate": 9.98651923370274e-06,
+      "loss": 0.5845,
+      "step": 542
+    },
+    {
+      "epoch": 0.05284671532846715,
+      "grad_norm": 1.4523428175637472,
+      "learning_rate": 9.986403323241252e-06,
+      "loss": 0.3817,
+      "step": 543
+    },
+    {
+      "epoch": 0.05294403892944039,
+      "grad_norm": 1.3085205057626972,
+      "learning_rate": 9.9862869172778e-06,
+      "loss": 0.294,
+      "step": 544
+    },
+    {
+      "epoch": 0.053041362530413624,
+      "grad_norm": 1.749260064779093,
+      "learning_rate": 9.986170015823953e-06,
+      "loss": 0.3885,
+      "step": 545
+    },
+    {
+      "epoch": 0.053138686131386864,
+      "grad_norm": 1.9224820302612053,
+      "learning_rate": 9.986052618891326e-06,
+      "loss": 0.5841,
+      "step": 546
+    },
+    {
+      "epoch": 0.0532360097323601,
+      "grad_norm": 1.6019594770490224,
+      "learning_rate": 9.985934726491587e-06,
+      "loss": 0.5602,
+      "step": 547
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 1.63788543125369,
+      "learning_rate": 9.98581633863645e-06,
+      "loss": 0.4913,
+      "step": 548
+    },
+    {
+      "epoch": 0.05343065693430657,
+      "grad_norm": 1.7751230304686407,
+      "learning_rate": 9.985697455337677e-06,
+      "loss": 0.4575,
+      "step": 549
+    },
+    {
+      "epoch": 0.0535279805352798,
+      "grad_norm": 1.4813830287768246,
+      "learning_rate": 9.985578076607086e-06,
+      "loss": 0.2811,
+      "step": 550
+    },
+    {
+      "epoch": 0.05362530413625304,
+      "grad_norm": 1.8047180833743464,
+      "learning_rate": 9.985458202456534e-06,
+      "loss": 0.5564,
+      "step": 551
+    },
+    {
+      "epoch": 0.053722627737226275,
+      "grad_norm": 1.4776771818705197,
+      "learning_rate": 9.985337832897938e-06,
+      "loss": 0.2842,
+      "step": 552
+    },
+    {
+      "epoch": 0.053819951338199515,
+      "grad_norm": 1.800973083472876,
+      "learning_rate": 9.985216967943256e-06,
+      "loss": 0.4017,
+      "step": 553
+    },
+    {
+      "epoch": 0.05391727493917275,
+      "grad_norm": 1.4167019147788764,
+      "learning_rate": 9.985095607604502e-06,
+      "loss": 0.2676,
+      "step": 554
+    },
+    {
+      "epoch": 0.05401459854014599,
+      "grad_norm": 1.462279330828973,
+      "learning_rate": 9.984973751893732e-06,
+      "loss": 0.342,
+      "step": 555
+    },
+    {
+      "epoch": 0.05411192214111922,
+      "grad_norm": 1.7941608662857766,
+      "learning_rate": 9.984851400823056e-06,
+      "loss": 0.4851,
+      "step": 556
+    },
+    {
+      "epoch": 0.05420924574209246,
+      "grad_norm": 1.865163176610701,
+      "learning_rate": 9.984728554404632e-06,
+      "loss": 0.5938,
+      "step": 557
+    },
+    {
+      "epoch": 0.05430656934306569,
+      "grad_norm": 1.9578700904261006,
+      "learning_rate": 9.984605212650669e-06,
+      "loss": 0.5846,
+      "step": 558
+    },
+    {
+      "epoch": 0.05440389294403893,
+      "grad_norm": 1.7615345522382602,
+      "learning_rate": 9.98448137557342e-06,
+      "loss": 0.5517,
+      "step": 559
+    },
+    {
+      "epoch": 0.054501216545012166,
+      "grad_norm": 1.7987507193579173,
+      "learning_rate": 9.984357043185195e-06,
+      "loss": 0.4511,
+      "step": 560
+    },
+    {
+      "epoch": 0.0545985401459854,
+      "grad_norm": 1.8966136067258859,
+      "learning_rate": 9.984232215498347e-06,
+      "loss": 0.3339,
+      "step": 561
+    },
+    {
+      "epoch": 0.05469586374695864,
+      "grad_norm": 1.760439118311743,
+      "learning_rate": 9.98410689252528e-06,
+      "loss": 0.4797,
+      "step": 562
+    },
+    {
+      "epoch": 0.05479318734793187,
+      "grad_norm": 1.7467534741216573,
+      "learning_rate": 9.983981074278448e-06,
+      "loss": 0.3854,
+      "step": 563
+    },
+    {
+      "epoch": 0.05489051094890511,
+      "grad_norm": 1.638747457914032,
+      "learning_rate": 9.983854760770353e-06,
+      "loss": 0.3215,
+      "step": 564
+    },
+    {
+      "epoch": 0.054987834549878344,
+      "grad_norm": 1.565721167011275,
+      "learning_rate": 9.983727952013546e-06,
+      "loss": 0.3573,
+      "step": 565
+    },
+    {
+      "epoch": 0.05508515815085158,
+      "grad_norm": 1.819373023432736,
+      "learning_rate": 9.98360064802063e-06,
+      "loss": 0.304,
+      "step": 566
+    },
+    {
+      "epoch": 0.055182481751824816,
+      "grad_norm": 2.219648367380945,
+      "learning_rate": 9.983472848804254e-06,
+      "loss": 0.7398,
+      "step": 567
+    },
+    {
+      "epoch": 0.055279805352798056,
+      "grad_norm": 1.7935096739228122,
+      "learning_rate": 9.98334455437712e-06,
+      "loss": 0.3257,
+      "step": 568
+    },
+    {
+      "epoch": 0.05537712895377129,
+      "grad_norm": 2.085379879601924,
+      "learning_rate": 9.983215764751971e-06,
+      "loss": 0.3477,
+      "step": 569
+    },
+    {
+      "epoch": 0.05547445255474453,
+      "grad_norm": 1.528881264990704,
+      "learning_rate": 9.98308647994161e-06,
+      "loss": 0.4173,
+      "step": 570
+    },
+    {
+      "epoch": 0.05557177615571776,
+      "grad_norm": 1.282510416609492,
+      "learning_rate": 9.982956699958883e-06,
+      "loss": 0.3513,
+      "step": 571
+    },
+    {
+      "epoch": 0.055669099756690994,
+      "grad_norm": 1.6035600811723405,
+      "learning_rate": 9.982826424816688e-06,
+      "loss": 0.3318,
+      "step": 572
+    },
+    {
+      "epoch": 0.055766423357664234,
+      "grad_norm": 1.9455996381881653,
+      "learning_rate": 9.982695654527966e-06,
+      "loss": 0.4991,
+      "step": 573
+    },
+    {
+      "epoch": 0.05586374695863747,
+      "grad_norm": 1.8397262762514839,
+      "learning_rate": 9.982564389105714e-06,
+      "loss": 0.345,
+      "step": 574
+    },
+    {
+      "epoch": 0.05596107055961071,
+      "grad_norm": 1.7997461351876956,
+      "learning_rate": 9.982432628562978e-06,
+      "loss": 0.5384,
+      "step": 575
+    },
+    {
+      "epoch": 0.05605839416058394,
+      "grad_norm": 1.6246101205121968,
+      "learning_rate": 9.982300372912848e-06,
+      "loss": 0.5499,
+      "step": 576
+    },
+    {
+      "epoch": 0.05615571776155718,
+      "grad_norm": 1.9184631207748861,
+      "learning_rate": 9.982167622168467e-06,
+      "loss": 0.449,
+      "step": 577
+    },
+    {
+      "epoch": 0.05625304136253041,
+      "grad_norm": 1.5368079698239796,
+      "learning_rate": 9.982034376343029e-06,
+      "loss": 0.3311,
+      "step": 578
+    },
+    {
+      "epoch": 0.05635036496350365,
+      "grad_norm": 1.9061539422519105,
+      "learning_rate": 9.98190063544977e-06,
+      "loss": 0.4182,
+      "step": 579
+    },
+    {
+      "epoch": 0.056447688564476885,
+      "grad_norm": 1.6727227174184238,
+      "learning_rate": 9.981766399501984e-06,
+      "loss": 0.482,
+      "step": 580
+    },
+    {
+      "epoch": 0.056545012165450124,
+      "grad_norm": 1.8546055763617424,
+      "learning_rate": 9.98163166851301e-06,
+      "loss": 0.5758,
+      "step": 581
+    },
+    {
+      "epoch": 0.05664233576642336,
+      "grad_norm": 2.0350303098403706,
+      "learning_rate": 9.981496442496234e-06,
+      "loss": 0.5236,
+      "step": 582
+    },
+    {
+      "epoch": 0.0567396593673966,
+      "grad_norm": 1.3907379790284926,
+      "learning_rate": 9.981360721465095e-06,
+      "loss": 0.3375,
+      "step": 583
+    },
+    {
+      "epoch": 0.05683698296836983,
+      "grad_norm": 2.0168702766261486,
+      "learning_rate": 9.98122450543308e-06,
+      "loss": 0.595,
+      "step": 584
+    },
+    {
+      "epoch": 0.05693430656934306,
+      "grad_norm": 1.7248754760467295,
+      "learning_rate": 9.981087794413722e-06,
+      "loss": 0.3747,
+      "step": 585
+    },
+    {
+      "epoch": 0.0570316301703163,
+      "grad_norm": 1.8918865818240052,
+      "learning_rate": 9.98095058842061e-06,
+      "loss": 0.5805,
+      "step": 586
+    },
+    {
+      "epoch": 0.057128953771289535,
+      "grad_norm": 1.8691153689026438,
+      "learning_rate": 9.980812887467377e-06,
+      "loss": 0.3451,
+      "step": 587
+    },
+    {
+      "epoch": 0.057226277372262775,
+      "grad_norm": 1.7475224395533677,
+      "learning_rate": 9.980674691567705e-06,
+      "loss": 0.2789,
+      "step": 588
+    },
+    {
+      "epoch": 0.05732360097323601,
+      "grad_norm": 1.876124489873064,
+      "learning_rate": 9.980536000735328e-06,
+      "loss": 0.5917,
+      "step": 589
+    },
+    {
+      "epoch": 0.05742092457420925,
+      "grad_norm": 1.6438847446693803,
+      "learning_rate": 9.980396814984025e-06,
+      "loss": 0.3063,
+      "step": 590
+    },
+    {
+      "epoch": 0.05751824817518248,
+      "grad_norm": 1.7609146888426583,
+      "learning_rate": 9.980257134327634e-06,
+      "loss": 0.4177,
+      "step": 591
+    },
+    {
+      "epoch": 0.05761557177615572,
+      "grad_norm": 3.1047413099950445,
+      "learning_rate": 9.980116958780027e-06,
+      "loss": 0.2793,
+      "step": 592
+    },
+    {
+      "epoch": 0.05771289537712895,
+      "grad_norm": 1.3365913263494138,
+      "learning_rate": 9.979976288355137e-06,
+      "loss": 0.2754,
+      "step": 593
+    },
+    {
+      "epoch": 0.05781021897810219,
+      "grad_norm": 1.7378721977452198,
+      "learning_rate": 9.979835123066943e-06,
+      "loss": 0.4156,
+      "step": 594
+    },
+    {
+      "epoch": 0.057907542579075426,
+      "grad_norm": 1.7652517953930271,
+      "learning_rate": 9.979693462929472e-06,
+      "loss": 0.3768,
+      "step": 595
+    },
+    {
+      "epoch": 0.05800486618004866,
+      "grad_norm": 2.4155692425963675,
+      "learning_rate": 9.979551307956801e-06,
+      "loss": 0.6409,
+      "step": 596
+    },
+    {
+      "epoch": 0.0581021897810219,
+      "grad_norm": 2.2339995809091913,
+      "learning_rate": 9.979408658163055e-06,
+      "loss": 0.3134,
+      "step": 597
+    },
+    {
+      "epoch": 0.05819951338199513,
+      "grad_norm": 1.9788468018769068,
+      "learning_rate": 9.97926551356241e-06,
+      "loss": 0.2509,
+      "step": 598
+    },
+    {
+      "epoch": 0.05829683698296837,
+      "grad_norm": 4.0668515887714385,
+      "learning_rate": 9.979121874169091e-06,
+      "loss": 0.3322,
+      "step": 599
+    },
+    {
+      "epoch": 0.058394160583941604,
+      "grad_norm": 2.0552497355613264,
+      "learning_rate": 9.97897773999737e-06,
+      "loss": 0.2732,
+      "step": 600
+    },
+    {
+      "epoch": 0.058491484184914844,
+      "grad_norm": 1.7372746328291984,
+      "learning_rate": 9.978833111061573e-06,
+      "loss": 0.3021,
+      "step": 601
+    },
+    {
+      "epoch": 0.058588807785888077,
+      "grad_norm": 1.8426989129405926,
+      "learning_rate": 9.978687987376067e-06,
+      "loss": 0.3147,
+      "step": 602
+    },
+    {
+      "epoch": 0.058686131386861316,
+      "grad_norm": 1.456816302033054,
+      "learning_rate": 9.978542368955278e-06,
+      "loss": 0.3669,
+      "step": 603
+    },
+    {
+      "epoch": 0.05878345498783455,
+      "grad_norm": 2.1398878847147973,
+      "learning_rate": 9.978396255813672e-06,
+      "loss": 0.457,
+      "step": 604
+    },
+    {
+      "epoch": 0.05888077858880779,
+      "grad_norm": 1.860652260495742,
+      "learning_rate": 9.978249647965769e-06,
+      "loss": 0.5567,
+      "step": 605
+    },
+    {
+      "epoch": 0.05897810218978102,
+      "grad_norm": 1.7559525207322428,
+      "learning_rate": 9.97810254542614e-06,
+      "loss": 0.439,
+      "step": 606
+    },
+    {
+      "epoch": 0.059075425790754255,
+      "grad_norm": 1.4912680944094816,
+      "learning_rate": 9.977954948209402e-06,
+      "loss": 0.4431,
+      "step": 607
+    },
+    {
+      "epoch": 0.059172749391727494,
+      "grad_norm": 1.766690700595448,
+      "learning_rate": 9.97780685633022e-06,
+      "loss": 0.3187,
+      "step": 608
+    },
+    {
+      "epoch": 0.05927007299270073,
+      "grad_norm": 2.169180646458804,
+      "learning_rate": 9.977658269803312e-06,
+      "loss": 0.5042,
+      "step": 609
+    },
+    {
+      "epoch": 0.05936739659367397,
+      "grad_norm": 1.623119439845207,
+      "learning_rate": 9.977509188643441e-06,
+      "loss": 0.3632,
+      "step": 610
+    },
+    {
+      "epoch": 0.0594647201946472,
+      "grad_norm": 2.0976883017366226,
+      "learning_rate": 9.977359612865424e-06,
+      "loss": 0.6465,
+      "step": 611
+    },
+    {
+      "epoch": 0.05956204379562044,
+      "grad_norm": 1.59126192242755,
+      "learning_rate": 9.977209542484123e-06,
+      "loss": 0.4335,
+      "step": 612
+    },
+    {
+      "epoch": 0.05965936739659367,
+      "grad_norm": 1.6532378246551842,
+      "learning_rate": 9.97705897751445e-06,
+      "loss": 0.3462,
+      "step": 613
+    },
+    {
+      "epoch": 0.05975669099756691,
+      "grad_norm": 1.6478059833585124,
+      "learning_rate": 9.976907917971365e-06,
+      "loss": 0.4063,
+      "step": 614
+    },
+    {
+      "epoch": 0.059854014598540145,
+      "grad_norm": 1.750559308727237,
+      "learning_rate": 9.976756363869884e-06,
+      "loss": 0.5062,
+      "step": 615
+    },
+    {
+      "epoch": 0.059951338199513385,
+      "grad_norm": 1.6400113365898012,
+      "learning_rate": 9.976604315225063e-06,
+      "loss": 0.3699,
+      "step": 616
+    },
+    {
+      "epoch": 0.06004866180048662,
+      "grad_norm": 1.5449283565959169,
+      "learning_rate": 9.976451772052013e-06,
+      "loss": 0.3635,
+      "step": 617
+    },
+    {
+      "epoch": 0.06014598540145986,
+      "grad_norm": 1.3799772345005799,
+      "learning_rate": 9.97629873436589e-06,
+      "loss": 0.2747,
+      "step": 618
+    },
+    {
+      "epoch": 0.06024330900243309,
+      "grad_norm": 1.9454941262632244,
+      "learning_rate": 9.976145202181905e-06,
+      "loss": 0.4963,
+      "step": 619
+    },
+    {
+      "epoch": 0.06034063260340632,
+      "grad_norm": 1.5274916477255973,
+      "learning_rate": 9.975991175515311e-06,
+      "loss": 0.3348,
+      "step": 620
+    },
+    {
+      "epoch": 0.06043795620437956,
+      "grad_norm": 1.9623540496009142,
+      "learning_rate": 9.975836654381416e-06,
+      "loss": 0.5373,
+      "step": 621
+    },
+    {
+      "epoch": 0.060535279805352796,
+      "grad_norm": 1.4248144765181632,
+      "learning_rate": 9.975681638795575e-06,
+      "loss": 0.3137,
+      "step": 622
+    },
+    {
+      "epoch": 0.060632603406326036,
+      "grad_norm": 1.4366236793713136,
+      "learning_rate": 9.975526128773192e-06,
+      "loss": 0.3519,
+      "step": 623
+    },
+    {
+      "epoch": 0.06072992700729927,
+      "grad_norm": 1.8458441140553945,
+      "learning_rate": 9.97537012432972e-06,
+      "loss": 0.3937,
+      "step": 624
+    },
+    {
+      "epoch": 0.06082725060827251,
+      "grad_norm": 1.868271580826056,
+      "learning_rate": 9.975213625480658e-06,
+      "loss": 0.4567,
+      "step": 625
+    },
+    {
+      "epoch": 0.06092457420924574,
+      "grad_norm": 2.4613001964869223,
+      "learning_rate": 9.97505663224156e-06,
+      "loss": 0.5607,
+      "step": 626
+    },
+    {
+      "epoch": 0.06102189781021898,
+      "grad_norm": 1.6709839772769468,
+      "learning_rate": 9.974899144628027e-06,
+      "loss": 0.3233,
+      "step": 627
+    },
+    {
+      "epoch": 0.061119221411192214,
+      "grad_norm": 1.8046591620263965,
+      "learning_rate": 9.97474116265571e-06,
+      "loss": 0.3929,
+      "step": 628
+    },
+    {
+      "epoch": 0.06121654501216545,
+      "grad_norm": 1.7182161369033975,
+      "learning_rate": 9.974582686340304e-06,
+      "loss": 0.3804,
+      "step": 629
+    },
+    {
+      "epoch": 0.061313868613138686,
+      "grad_norm": 2.435940855169524,
+      "learning_rate": 9.974423715697558e-06,
+      "loss": 0.7453,
+      "step": 630
+    },
+    {
+      "epoch": 0.06141119221411192,
+      "grad_norm": 1.401143104634322,
+      "learning_rate": 9.974264250743272e-06,
+      "loss": 0.306,
+      "step": 631
+    },
+    {
+      "epoch": 0.06150851581508516,
+      "grad_norm": 1.540550326071636,
+      "learning_rate": 9.97410429149329e-06,
+      "loss": 0.3582,
+      "step": 632
+    },
+    {
+      "epoch": 0.06160583941605839,
+      "grad_norm": 4.038520112503673,
+      "learning_rate": 9.973943837963507e-06,
+      "loss": 0.2688,
+      "step": 633
+    },
+    {
+      "epoch": 0.06170316301703163,
+      "grad_norm": 2.032927304778425,
+      "learning_rate": 9.973782890169867e-06,
+      "loss": 0.6952,
+      "step": 634
+    },
+    {
+      "epoch": 0.061800486618004864,
+      "grad_norm": 1.5242884680104736,
+      "learning_rate": 9.973621448128364e-06,
+      "loss": 0.3957,
+      "step": 635
+    },
+    {
+      "epoch": 0.061897810218978104,
+      "grad_norm": 1.599953340803732,
+      "learning_rate": 9.973459511855042e-06,
+      "loss": 0.3783,
+      "step": 636
+    },
+    {
+      "epoch": 0.06199513381995134,
+      "grad_norm": 2.1886899708740697,
+      "learning_rate": 9.973297081365988e-06,
+      "loss": 0.5426,
+      "step": 637
+    },
+    {
+      "epoch": 0.06209245742092458,
+      "grad_norm": 1.363421719809718,
+      "learning_rate": 9.973134156677349e-06,
+      "loss": 0.2707,
+      "step": 638
+    },
+    {
+      "epoch": 0.06218978102189781,
+      "grad_norm": 1.883218491971664,
+      "learning_rate": 9.972970737805312e-06,
+      "loss": 0.543,
+      "step": 639
+    },
+    {
+      "epoch": 0.06228710462287105,
+      "grad_norm": 1.6336178778276322,
+      "learning_rate": 9.972806824766117e-06,
+      "loss": 0.4833,
+      "step": 640
+    },
+    {
+      "epoch": 0.06238442822384428,
+      "grad_norm": 1.74145478719615,
+      "learning_rate": 9.972642417576049e-06,
+      "loss": 0.5456,
+      "step": 641
+    },
+    {
+      "epoch": 0.062481751824817515,
+      "grad_norm": 1.3939447959630629,
+      "learning_rate": 9.972477516251448e-06,
+      "loss": 0.2935,
+      "step": 642
+    },
+    {
+      "epoch": 0.06257907542579075,
+      "grad_norm": 1.9741261661680443,
+      "learning_rate": 9.9723121208087e-06,
+      "loss": 0.4377,
+      "step": 643
+    },
+    {
+      "epoch": 0.06267639902676399,
+      "grad_norm": 2.214700253529172,
+      "learning_rate": 9.972146231264242e-06,
+      "loss": 0.6711,
+      "step": 644
+    },
+    {
+      "epoch": 0.06277372262773723,
+      "grad_norm": 1.7399845992974294,
+      "learning_rate": 9.971979847634554e-06,
+      "loss": 0.5327,
+      "step": 645
+    },
+    {
+      "epoch": 0.06287104622871047,
+      "grad_norm": 1.3552365502663122,
+      "learning_rate": 9.971812969936174e-06,
+      "loss": 0.3553,
+      "step": 646
+    },
+    {
+      "epoch": 0.06296836982968369,
+      "grad_norm": 1.8378075997453163,
+      "learning_rate": 9.971645598185685e-06,
+      "loss": 0.3709,
+      "step": 647
+    },
+    {
+      "epoch": 0.06306569343065693,
+      "grad_norm": 1.7441350204189767,
+      "learning_rate": 9.971477732399714e-06,
+      "loss": 0.489,
+      "step": 648
+    },
+    {
+      "epoch": 0.06316301703163017,
+      "grad_norm": 2.083031963167252,
+      "learning_rate": 9.971309372594947e-06,
+      "loss": 0.6196,
+      "step": 649
+    },
+    {
+      "epoch": 0.06326034063260341,
+      "grad_norm": 1.5678236487001533,
+      "learning_rate": 9.971140518788112e-06,
+      "loss": 0.3202,
+      "step": 650
+    },
+    {
+      "epoch": 0.06335766423357664,
+      "grad_norm": 1.7281008810115812,
+      "learning_rate": 9.970971170995988e-06,
+      "loss": 0.4169,
+      "step": 651
+    },
+    {
+      "epoch": 0.06345498783454988,
+      "grad_norm": 1.5626981990993993,
+      "learning_rate": 9.970801329235402e-06,
+      "loss": 0.4238,
+      "step": 652
+    },
+    {
+      "epoch": 0.06355231143552312,
+      "grad_norm": 1.5338214380715702,
+      "learning_rate": 9.970630993523234e-06,
+      "loss": 0.278,
+      "step": 653
+    },
+    {
+      "epoch": 0.06364963503649634,
+      "grad_norm": 1.7806299033721755,
+      "learning_rate": 9.970460163876409e-06,
+      "loss": 0.5649,
+      "step": 654
+    },
+    {
+      "epoch": 0.06374695863746958,
+      "grad_norm": 1.9349681554929028,
+      "learning_rate": 9.9702888403119e-06,
+      "loss": 0.3297,
+      "step": 655
+    },
+    {
+      "epoch": 0.06384428223844282,
+      "grad_norm": 1.4947723050696704,
+      "learning_rate": 9.970117022846736e-06,
+      "loss": 0.4077,
+      "step": 656
+    },
+    {
+      "epoch": 0.06394160583941606,
+      "grad_norm": 1.5696774237596223,
+      "learning_rate": 9.96994471149799e-06,
+      "loss": 0.4681,
+      "step": 657
+    },
+    {
+      "epoch": 0.06403892944038929,
+      "grad_norm": 1.7662095984112474,
+      "learning_rate": 9.969771906282781e-06,
+      "loss": 0.539,
+      "step": 658
+    },
+    {
+      "epoch": 0.06413625304136253,
+      "grad_norm": 2.926336951253308,
+      "learning_rate": 9.969598607218285e-06,
+      "loss": 0.4196,
+      "step": 659
+    },
+    {
+      "epoch": 0.06423357664233577,
+      "grad_norm": 3.148192138198314,
+      "learning_rate": 9.96942481432172e-06,
+      "loss": 0.4827,
+      "step": 660
+    },
+    {
+      "epoch": 0.06433090024330901,
+      "grad_norm": 1.790436662552377,
+      "learning_rate": 9.969250527610356e-06,
+      "loss": 0.4972,
+      "step": 661
+    },
+    {
+      "epoch": 0.06442822384428223,
+      "grad_norm": 1.4712739725679773,
+      "learning_rate": 9.969075747101514e-06,
+      "loss": 0.4112,
+      "step": 662
+    },
+    {
+      "epoch": 0.06452554744525547,
+      "grad_norm": 1.4521996617982842,
+      "learning_rate": 9.96890047281256e-06,
+      "loss": 0.3729,
+      "step": 663
+    },
+    {
+      "epoch": 0.06462287104622871,
+      "grad_norm": 1.5457088814513262,
+      "learning_rate": 9.96872470476091e-06,
+      "loss": 0.4294,
+      "step": 664
+    },
+    {
+      "epoch": 0.06472019464720194,
+      "grad_norm": 1.7644033340951866,
+      "learning_rate": 9.968548442964034e-06,
+      "loss": 0.4487,
+      "step": 665
+    },
+    {
+      "epoch": 0.06481751824817518,
+      "grad_norm": 1.632555708701406,
+      "learning_rate": 9.968371687439446e-06,
+      "loss": 0.3929,
+      "step": 666
+    },
+    {
+      "epoch": 0.06491484184914842,
+      "grad_norm": 1.8990302396780172,
+      "learning_rate": 9.968194438204708e-06,
+      "loss": 0.4101,
+      "step": 667
+    },
+    {
+      "epoch": 0.06501216545012166,
+      "grad_norm": 2.092762728551112,
+      "learning_rate": 9.968016695277436e-06,
+      "loss": 0.5712,
+      "step": 668
+    },
+    {
+      "epoch": 0.06510948905109488,
+      "grad_norm": 1.5876668887386824,
+      "learning_rate": 9.967838458675292e-06,
+      "loss": 0.494,
+      "step": 669
+    },
+    {
+      "epoch": 0.06520681265206812,
+      "grad_norm": 1.7536517597940893,
+      "learning_rate": 9.967659728415985e-06,
+      "loss": 0.6121,
+      "step": 670
+    },
+    {
+      "epoch": 0.06530413625304136,
+      "grad_norm": 1.9021294255711243,
+      "learning_rate": 9.96748050451728e-06,
+      "loss": 0.3634,
+      "step": 671
+    },
+    {
+      "epoch": 0.0654014598540146,
+      "grad_norm": 1.4457078547633553,
+      "learning_rate": 9.96730078699698e-06,
+      "loss": 0.4586,
+      "step": 672
+    },
+    {
+      "epoch": 0.06549878345498783,
+      "grad_norm": 1.6474950184261972,
+      "learning_rate": 9.967120575872952e-06,
+      "loss": 0.5028,
+      "step": 673
+    },
+    {
+      "epoch": 0.06559610705596107,
+      "grad_norm": 1.9901979572232373,
+      "learning_rate": 9.966939871163098e-06,
+      "loss": 0.6986,
+      "step": 674
+    },
+    {
+      "epoch": 0.06569343065693431,
+      "grad_norm": 1.3671458210722949,
+      "learning_rate": 9.966758672885375e-06,
+      "loss": 0.3945,
+      "step": 675
+    },
+    {
+      "epoch": 0.06579075425790755,
+      "grad_norm": 1.8371332697903162,
+      "learning_rate": 9.96657698105779e-06,
+      "loss": 0.6782,
+      "step": 676
+    },
+    {
+      "epoch": 0.06588807785888078,
+      "grad_norm": 1.1955013749239556,
+      "learning_rate": 9.966394795698397e-06,
+      "loss": 0.242,
+      "step": 677
+    },
+    {
+      "epoch": 0.06598540145985402,
+      "grad_norm": 1.5330975344313047,
+      "learning_rate": 9.966212116825302e-06,
+      "loss": 0.4351,
+      "step": 678
+    },
+    {
+      "epoch": 0.06608272506082725,
+      "grad_norm": 1.539581985713935,
+      "learning_rate": 9.966028944456657e-06,
+      "loss": 0.3512,
+      "step": 679
+    },
+    {
+      "epoch": 0.06618004866180048,
+      "grad_norm": 1.9573455375443363,
+      "learning_rate": 9.965845278610661e-06,
+      "loss": 0.4859,
+      "step": 680
+    },
+    {
+      "epoch": 0.06627737226277372,
+      "grad_norm": 1.8387055004344444,
+      "learning_rate": 9.96566111930557e-06,
+      "loss": 0.3831,
+      "step": 681
+    },
+    {
+      "epoch": 0.06637469586374696,
+      "grad_norm": 1.7056154014174738,
+      "learning_rate": 9.96547646655968e-06,
+      "loss": 0.4675,
+      "step": 682
+    },
+    {
+      "epoch": 0.0664720194647202,
+      "grad_norm": 1.881602931580563,
+      "learning_rate": 9.965291320391342e-06,
+      "loss": 0.5955,
+      "step": 683
+    },
+    {
+      "epoch": 0.06656934306569343,
+      "grad_norm": 2.9885065529853416,
+      "learning_rate": 9.965105680818955e-06,
+      "loss": 0.393,
+      "step": 684
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 1.7363492709096229,
+      "learning_rate": 9.964919547860963e-06,
+      "loss": 0.4903,
+      "step": 685
+    },
+    {
+      "epoch": 0.0667639902676399,
+      "grad_norm": 1.8182376939684146,
+      "learning_rate": 9.964732921535863e-06,
+      "loss": 0.5443,
+      "step": 686
+    },
+    {
+      "epoch": 0.06686131386861315,
+      "grad_norm": 1.6914779965026407,
+      "learning_rate": 9.964545801862202e-06,
+      "loss": 0.5119,
+      "step": 687
+    },
+    {
+      "epoch": 0.06695863746958637,
+      "grad_norm": 1.2736843314571082,
+      "learning_rate": 9.964358188858573e-06,
+      "loss": 0.2495,
+      "step": 688
+    },
+    {
+      "epoch": 0.06705596107055961,
+      "grad_norm": 1.5831736266585599,
+      "learning_rate": 9.96417008254362e-06,
+      "loss": 0.4489,
+      "step": 689
+    },
+    {
+      "epoch": 0.06715328467153285,
+      "grad_norm": 2.2148297560046224,
+      "learning_rate": 9.963981482936034e-06,
+      "loss": 0.5415,
+      "step": 690
+    },
+    {
+      "epoch": 0.06725060827250608,
+      "grad_norm": 1.5025934211262992,
+      "learning_rate": 9.963792390054558e-06,
+      "loss": 0.3903,
+      "step": 691
+    },
+    {
+      "epoch": 0.06734793187347932,
+      "grad_norm": 1.4602374679322867,
+      "learning_rate": 9.96360280391798e-06,
+      "loss": 0.3199,
+      "step": 692
+    },
+    {
+      "epoch": 0.06744525547445256,
+      "grad_norm": 1.5813416284844282,
+      "learning_rate": 9.963412724545142e-06,
+      "loss": 0.3213,
+      "step": 693
+    },
+    {
+      "epoch": 0.0675425790754258,
+      "grad_norm": 1.246883512769049,
+      "learning_rate": 9.96322215195493e-06,
+      "loss": 0.2644,
+      "step": 694
+    },
+    {
+      "epoch": 0.06763990267639902,
+      "grad_norm": 1.7094335347253355,
+      "learning_rate": 9.963031086166282e-06,
+      "loss": 0.4761,
+      "step": 695
+    },
+    {
+      "epoch": 0.06773722627737226,
+      "grad_norm": 1.6516611118524773,
+      "learning_rate": 9.962839527198184e-06,
+      "loss": 0.4823,
+      "step": 696
+    },
+    {
+      "epoch": 0.0678345498783455,
+      "grad_norm": 1.3531669839243998,
+      "learning_rate": 9.962647475069672e-06,
+      "loss": 0.4272,
+      "step": 697
+    },
+    {
+      "epoch": 0.06793187347931874,
+      "grad_norm": 1.9430916606586504,
+      "learning_rate": 9.962454929799829e-06,
+      "loss": 0.5776,
+      "step": 698
+    },
+    {
+      "epoch": 0.06802919708029197,
+      "grad_norm": 1.8772536403383466,
+      "learning_rate": 9.962261891407792e-06,
+      "loss": 0.6338,
+      "step": 699
+    },
+    {
+      "epoch": 0.0681265206812652,
+      "grad_norm": 1.3972932620324034,
+      "learning_rate": 9.96206835991274e-06,
+      "loss": 0.3671,
+      "step": 700
+    },
+    {
+      "epoch": 0.06822384428223845,
+      "grad_norm": 1.287329601381866,
+      "learning_rate": 9.961874335333904e-06,
+      "loss": 0.2744,
+      "step": 701
+    },
+    {
+      "epoch": 0.06832116788321167,
+      "grad_norm": 1.5600519457751545,
+      "learning_rate": 9.961679817690566e-06,
+      "loss": 0.4433,
+      "step": 702
+    },
+    {
+      "epoch": 0.06841849148418491,
+      "grad_norm": 1.3898736874388666,
+      "learning_rate": 9.961484807002056e-06,
+      "loss": 0.4197,
+      "step": 703
+    },
+    {
+      "epoch": 0.06851581508515815,
+      "grad_norm": 1.672202746628868,
+      "learning_rate": 9.961289303287749e-06,
+      "loss": 0.4601,
+      "step": 704
+    },
+    {
+      "epoch": 0.06861313868613139,
+      "grad_norm": 1.7427655274680753,
+      "learning_rate": 9.961093306567076e-06,
+      "loss": 0.5845,
+      "step": 705
+    },
+    {
+      "epoch": 0.06871046228710462,
+      "grad_norm": 1.794570108008766,
+      "learning_rate": 9.960896816859512e-06,
+      "loss": 0.3459,
+      "step": 706
+    },
+    {
+      "epoch": 0.06880778588807786,
+      "grad_norm": 1.6024314197975584,
+      "learning_rate": 9.960699834184582e-06,
+      "loss": 0.4441,
+      "step": 707
+    },
+    {
+      "epoch": 0.0689051094890511,
+      "grad_norm": 1.619306935418848,
+      "learning_rate": 9.960502358561858e-06,
+      "loss": 0.4647,
+      "step": 708
+    },
+    {
+      "epoch": 0.06900243309002434,
+      "grad_norm": 1.5009190604836247,
+      "learning_rate": 9.960304390010968e-06,
+      "loss": 0.373,
+      "step": 709
+    },
+    {
+      "epoch": 0.06909975669099756,
+      "grad_norm": 1.8613999824223078,
+      "learning_rate": 9.960105928551583e-06,
+      "loss": 0.3926,
+      "step": 710
+    },
+    {
+      "epoch": 0.0691970802919708,
+      "grad_norm": 2.8907340364253757,
+      "learning_rate": 9.959906974203422e-06,
+      "loss": 0.5451,
+      "step": 711
+    },
+    {
+      "epoch": 0.06929440389294404,
+      "grad_norm": 1.826374356881247,
+      "learning_rate": 9.959707526986256e-06,
+      "loss": 0.4341,
+      "step": 712
+    },
+    {
+      "epoch": 0.06939172749391727,
+      "grad_norm": 2.5001373253299133,
+      "learning_rate": 9.959507586919903e-06,
+      "loss": 0.6643,
+      "step": 713
+    },
+    {
+      "epoch": 0.06948905109489051,
+      "grad_norm": 1.769427365923108,
+      "learning_rate": 9.959307154024234e-06,
+      "loss": 0.5431,
+      "step": 714
+    },
+    {
+      "epoch": 0.06958637469586375,
+      "grad_norm": 2.3285358245695322,
+      "learning_rate": 9.959106228319166e-06,
+      "loss": 0.5274,
+      "step": 715
+    },
+    {
+      "epoch": 0.06968369829683699,
+      "grad_norm": 1.4070234926508725,
+      "learning_rate": 9.958904809824663e-06,
+      "loss": 0.3257,
+      "step": 716
+    },
+    {
+      "epoch": 0.06978102189781021,
+      "grad_norm": 1.9284568290872997,
+      "learning_rate": 9.958702898560742e-06,
+      "loss": 0.5648,
+      "step": 717
+    },
+    {
+      "epoch": 0.06987834549878345,
+      "grad_norm": 2.092543866644565,
+      "learning_rate": 9.958500494547465e-06,
+      "loss": 0.6256,
+      "step": 718
+    },
+    {
+      "epoch": 0.0699756690997567,
+      "grad_norm": 1.5948763588365042,
+      "learning_rate": 9.958297597804947e-06,
+      "loss": 0.4011,
+      "step": 719
+    },
+    {
+      "epoch": 0.07007299270072993,
+      "grad_norm": 1.2246362905267065,
+      "learning_rate": 9.958094208353348e-06,
+      "loss": 0.2444,
+      "step": 720
+    },
+    {
+      "epoch": 0.07017031630170316,
+      "grad_norm": 1.2302916868666773,
+      "learning_rate": 9.95789032621288e-06,
+      "loss": 0.3191,
+      "step": 721
+    },
+    {
+      "epoch": 0.0702676399026764,
+      "grad_norm": 1.5504396768673763,
+      "learning_rate": 9.957685951403803e-06,
+      "loss": 0.3112,
+      "step": 722
+    },
+    {
+      "epoch": 0.07036496350364964,
+      "grad_norm": 2.1205819146422438,
+      "learning_rate": 9.957481083946427e-06,
+      "loss": 0.3453,
+      "step": 723
+    },
+    {
+      "epoch": 0.07046228710462286,
+      "grad_norm": 2.048519725880563,
+      "learning_rate": 9.957275723861108e-06,
+      "loss": 0.5266,
+      "step": 724
+    },
+    {
+      "epoch": 0.0705596107055961,
+      "grad_norm": 1.4453693275620771,
+      "learning_rate": 9.957069871168253e-06,
+      "loss": 0.3082,
+      "step": 725
+    },
+    {
+      "epoch": 0.07065693430656934,
+      "grad_norm": 1.8824931146868138,
+      "learning_rate": 9.956863525888318e-06,
+      "loss": 0.588,
+      "step": 726
+    },
+    {
+      "epoch": 0.07075425790754258,
+      "grad_norm": 1.6143333569692804,
+      "learning_rate": 9.956656688041807e-06,
+      "loss": 0.4126,
+      "step": 727
+    },
+    {
+      "epoch": 0.07085158150851581,
+      "grad_norm": 1.7905307392122496,
+      "learning_rate": 9.956449357649276e-06,
+      "loss": 0.521,
+      "step": 728
+    },
+    {
+      "epoch": 0.07094890510948905,
+      "grad_norm": 1.3295021098228834,
+      "learning_rate": 9.956241534731325e-06,
+      "loss": 0.31,
+      "step": 729
+    },
+    {
+      "epoch": 0.07104622871046229,
+      "grad_norm": 1.5783278835300563,
+      "learning_rate": 9.956033219308607e-06,
+      "loss": 0.3091,
+      "step": 730
+    },
+    {
+      "epoch": 0.07114355231143553,
+      "grad_norm": 1.9905003004076265,
+      "learning_rate": 9.955824411401822e-06,
+      "loss": 0.3843,
+      "step": 731
+    },
+    {
+      "epoch": 0.07124087591240875,
+      "grad_norm": 1.7644558301646922,
+      "learning_rate": 9.955615111031717e-06,
+      "loss": 0.4288,
+      "step": 732
+    },
+    {
+      "epoch": 0.071338199513382,
+      "grad_norm": 1.5922207695027908,
+      "learning_rate": 9.955405318219096e-06,
+      "loss": 0.4767,
+      "step": 733
+    },
+    {
+      "epoch": 0.07143552311435523,
+      "grad_norm": 1.7054240956141933,
+      "learning_rate": 9.955195032984798e-06,
+      "loss": 0.4082,
+      "step": 734
+    },
+    {
+      "epoch": 0.07153284671532846,
+      "grad_norm": 1.3954970063738148,
+      "learning_rate": 9.954984255349729e-06,
+      "loss": 0.318,
+      "step": 735
+    },
+    {
+      "epoch": 0.0716301703163017,
+      "grad_norm": 1.7287069268697828,
+      "learning_rate": 9.954772985334825e-06,
+      "loss": 0.4998,
+      "step": 736
+    },
+    {
+      "epoch": 0.07172749391727494,
+      "grad_norm": 1.4535895804720915,
+      "learning_rate": 9.954561222961086e-06,
+      "loss": 0.2489,
+      "step": 737
+    },
+    {
+      "epoch": 0.07182481751824818,
+      "grad_norm": 1.7113518757446542,
+      "learning_rate": 9.954348968249552e-06,
+      "loss": 0.4578,
+      "step": 738
+    },
+    {
+      "epoch": 0.0719221411192214,
+      "grad_norm": 1.6741613993254088,
+      "learning_rate": 9.954136221221316e-06,
+      "loss": 0.4907,
+      "step": 739
+    },
+    {
+      "epoch": 0.07201946472019465,
+      "grad_norm": 1.590982465166657,
+      "learning_rate": 9.95392298189752e-06,
+      "loss": 0.4116,
+      "step": 740
+    },
+    {
+      "epoch": 0.07211678832116789,
+      "grad_norm": 1.422974716648181,
+      "learning_rate": 9.953709250299351e-06,
+      "loss": 0.3501,
+      "step": 741
+    },
+    {
+      "epoch": 0.07221411192214112,
+      "grad_norm": 1.8424007198547667,
+      "learning_rate": 9.953495026448048e-06,
+      "loss": 0.5647,
+      "step": 742
+    },
+    {
+      "epoch": 0.07231143552311435,
+      "grad_norm": 1.6572484299897867,
+      "learning_rate": 9.953280310364902e-06,
+      "loss": 0.3937,
+      "step": 743
+    },
+    {
+      "epoch": 0.07240875912408759,
+      "grad_norm": 1.6027770112754065,
+      "learning_rate": 9.953065102071245e-06,
+      "loss": 0.3845,
+      "step": 744
+    },
+    {
+      "epoch": 0.07250608272506083,
+      "grad_norm": 1.3618658637431431,
+      "learning_rate": 9.952849401588464e-06,
+      "loss": 0.3946,
+      "step": 745
+    },
+    {
+      "epoch": 0.07260340632603407,
+      "grad_norm": 1.63075572158439,
+      "learning_rate": 9.952633208937997e-06,
+      "loss": 0.4506,
+      "step": 746
+    },
+    {
+      "epoch": 0.0727007299270073,
+      "grad_norm": 1.483187632244976,
+      "learning_rate": 9.95241652414132e-06,
+      "loss": 0.3908,
+      "step": 747
+    },
+    {
+      "epoch": 0.07279805352798054,
+      "grad_norm": 2.147960263046311,
+      "learning_rate": 9.952199347219972e-06,
+      "loss": 0.5249,
+      "step": 748
+    },
+    {
+      "epoch": 0.07289537712895378,
+      "grad_norm": 1.5046941105429004,
+      "learning_rate": 9.951981678195529e-06,
+      "loss": 0.3592,
+      "step": 749
+    },
+    {
+      "epoch": 0.072992700729927,
+      "grad_norm": 1.1457618113072725,
+      "learning_rate": 9.951763517089624e-06,
+      "loss": 0.2197,
+      "step": 750
+    },
+    {
+      "epoch": 0.07309002433090024,
+      "grad_norm": 1.9275946136488011,
+      "learning_rate": 9.951544863923934e-06,
+      "loss": 0.5692,
+      "step": 751
+    },
+    {
+      "epoch": 0.07318734793187348,
+      "grad_norm": 1.9590929330277462,
+      "learning_rate": 9.95132571872019e-06,
+      "loss": 0.7243,
+      "step": 752
+    },
+    {
+      "epoch": 0.07328467153284672,
+      "grad_norm": 2.1368780826391283,
+      "learning_rate": 9.951106081500162e-06,
+      "loss": 0.7601,
+      "step": 753
+    },
+    {
+      "epoch": 0.07338199513381995,
+      "grad_norm": 2.0085695969306396,
+      "learning_rate": 9.950885952285682e-06,
+      "loss": 0.5541,
+      "step": 754
+    },
+    {
+      "epoch": 0.07347931873479319,
+      "grad_norm": 1.9283983503616706,
+      "learning_rate": 9.950665331098622e-06,
+      "loss": 0.3832,
+      "step": 755
+    },
+    {
+      "epoch": 0.07357664233576643,
+      "grad_norm": 1.4173732379297153,
+      "learning_rate": 9.950444217960902e-06,
+      "loss": 0.379,
+      "step": 756
+    },
+    {
+      "epoch": 0.07367396593673967,
+      "grad_norm": 1.5015176407129935,
+      "learning_rate": 9.9502226128945e-06,
+      "loss": 0.4696,
+      "step": 757
+    },
+    {
+      "epoch": 0.07377128953771289,
+      "grad_norm": 1.6746905852394565,
+      "learning_rate": 9.950000515921434e-06,
+      "loss": 0.2984,
+      "step": 758
+    },
+    {
+      "epoch": 0.07386861313868613,
+      "grad_norm": 1.4429847737048944,
+      "learning_rate": 9.949777927063776e-06,
+      "loss": 0.3748,
+      "step": 759
+    },
+    {
+      "epoch": 0.07396593673965937,
+      "grad_norm": 1.1895632638034424,
+      "learning_rate": 9.94955484634364e-06,
+      "loss": 0.3014,
+      "step": 760
+    },
+    {
+      "epoch": 0.0740632603406326,
+      "grad_norm": 1.5497241513071458,
+      "learning_rate": 9.949331273783198e-06,
+      "loss": 0.5458,
+      "step": 761
+    },
+    {
+      "epoch": 0.07416058394160584,
+      "grad_norm": 1.5531214201672936,
+      "learning_rate": 9.949107209404664e-06,
+      "loss": 0.4575,
+      "step": 762
+    },
+    {
+      "epoch": 0.07425790754257908,
+      "grad_norm": 1.3336107839559097,
+      "learning_rate": 9.948882653230306e-06,
+      "loss": 0.4227,
+      "step": 763
+    },
+    {
+      "epoch": 0.07435523114355232,
+      "grad_norm": 1.7418209768074853,
+      "learning_rate": 9.948657605282437e-06,
+      "loss": 0.659,
+      "step": 764
+    },
+    {
+      "epoch": 0.07445255474452554,
+      "grad_norm": 1.462439433090815,
+      "learning_rate": 9.94843206558342e-06,
+      "loss": 0.445,
+      "step": 765
+    },
+    {
+      "epoch": 0.07454987834549878,
+      "grad_norm": 1.0856086178050317,
+      "learning_rate": 9.948206034155666e-06,
+      "loss": 0.2245,
+      "step": 766
+    },
+    {
+      "epoch": 0.07464720194647202,
+      "grad_norm": 1.458503858496447,
+      "learning_rate": 9.947979511021638e-06,
+      "loss": 0.3009,
+      "step": 767
+    },
+    {
+      "epoch": 0.07474452554744526,
+      "grad_norm": 1.1921292471996519,
+      "learning_rate": 9.947752496203844e-06,
+      "loss": 0.2988,
+      "step": 768
+    },
+    {
+      "epoch": 0.07484184914841849,
+      "grad_norm": 1.6693024138876786,
+      "learning_rate": 9.947524989724844e-06,
+      "loss": 0.4783,
+      "step": 769
+    },
+    {
+      "epoch": 0.07493917274939173,
+      "grad_norm": 1.4928671202909605,
+      "learning_rate": 9.947296991607244e-06,
+      "loss": 0.4161,
+      "step": 770
+    },
+    {
+      "epoch": 0.07503649635036497,
+      "grad_norm": 1.4549005796935413,
+      "learning_rate": 9.947068501873702e-06,
+      "loss": 0.4186,
+      "step": 771
+    },
+    {
+      "epoch": 0.0751338199513382,
+      "grad_norm": 1.7544781744298734,
+      "learning_rate": 9.946839520546923e-06,
+      "loss": 0.5593,
+      "step": 772
+    },
+    {
+      "epoch": 0.07523114355231143,
+      "grad_norm": 1.561541454027553,
+      "learning_rate": 9.946610047649659e-06,
+      "loss": 0.5097,
+      "step": 773
+    },
+    {
+      "epoch": 0.07532846715328467,
+      "grad_norm": 1.598616630831168,
+      "learning_rate": 9.946380083204714e-06,
+      "loss": 0.3744,
+      "step": 774
+    },
+    {
+      "epoch": 0.07542579075425791,
+      "grad_norm": 1.6915556597188157,
+      "learning_rate": 9.94614962723494e-06,
+      "loss": 0.439,
+      "step": 775
+    },
+    {
+      "epoch": 0.07552311435523114,
+      "grad_norm": 1.220024420697048,
+      "learning_rate": 9.945918679763237e-06,
+      "loss": 0.2339,
+      "step": 776
+    },
+    {
+      "epoch": 0.07562043795620438,
+      "grad_norm": 1.6061445238682988,
+      "learning_rate": 9.945687240812556e-06,
+      "loss": 0.4493,
+      "step": 777
+    },
+    {
+      "epoch": 0.07571776155717762,
+      "grad_norm": 1.400813806243779,
+      "learning_rate": 9.945455310405895e-06,
+      "loss": 0.4513,
+      "step": 778
+    },
+    {
+      "epoch": 0.07581508515815086,
+      "grad_norm": 1.753751480308555,
+      "learning_rate": 9.945222888566298e-06,
+      "loss": 0.5379,
+      "step": 779
+    },
+    {
+      "epoch": 0.07591240875912408,
+      "grad_norm": 1.4421667558329163,
+      "learning_rate": 9.944989975316862e-06,
+      "loss": 0.4118,
+      "step": 780
+    },
+    {
+      "epoch": 0.07600973236009732,
+      "grad_norm": 1.4411974086247974,
+      "learning_rate": 9.944756570680733e-06,
+      "loss": 0.3295,
+      "step": 781
+    },
+    {
+      "epoch": 0.07610705596107056,
+      "grad_norm": 1.5545586767450623,
+      "learning_rate": 9.944522674681107e-06,
+      "loss": 0.4146,
+      "step": 782
+    },
+    {
+      "epoch": 0.07620437956204379,
+      "grad_norm": 2.0019900434858084,
+      "learning_rate": 9.944288287341222e-06,
+      "loss": 0.4945,
+      "step": 783
+    },
+    {
+      "epoch": 0.07630170316301703,
+      "grad_norm": 1.5834930071710975,
+      "learning_rate": 9.944053408684371e-06,
+      "loss": 0.3781,
+      "step": 784
+    },
+    {
+      "epoch": 0.07639902676399027,
+      "grad_norm": 1.5272521164667598,
+      "learning_rate": 9.943818038733894e-06,
+      "loss": 0.3865,
+      "step": 785
+    },
+    {
+      "epoch": 0.07649635036496351,
+      "grad_norm": 1.8005925077547513,
+      "learning_rate": 9.94358217751318e-06,
+      "loss": 0.3951,
+      "step": 786
+    },
+    {
+      "epoch": 0.07659367396593673,
+      "grad_norm": 2.0471085276865995,
+      "learning_rate": 9.943345825045664e-06,
+      "loss": 0.6391,
+      "step": 787
+    },
+    {
+      "epoch": 0.07669099756690997,
+      "grad_norm": 1.7893386028077656,
+      "learning_rate": 9.943108981354839e-06,
+      "loss": 0.6373,
+      "step": 788
+    },
+    {
+      "epoch": 0.07678832116788321,
+      "grad_norm": 1.6529186502183046,
+      "learning_rate": 9.942871646464234e-06,
+      "loss": 0.4901,
+      "step": 789
+    },
+    {
+      "epoch": 0.07688564476885645,
+      "grad_norm": 1.8449837387732961,
+      "learning_rate": 9.942633820397436e-06,
+      "loss": 0.4444,
+      "step": 790
+    },
+    {
+      "epoch": 0.07698296836982968,
+      "grad_norm": 1.5278738521461448,
+      "learning_rate": 9.942395503178077e-06,
+      "loss": 0.3701,
+      "step": 791
+    },
+    {
+      "epoch": 0.07708029197080292,
+      "grad_norm": 1.8197808533034088,
+      "learning_rate": 9.942156694829838e-06,
+      "loss": 0.6142,
+      "step": 792
+    },
+    {
+      "epoch": 0.07717761557177616,
+      "grad_norm": 1.8496691201700692,
+      "learning_rate": 9.941917395376452e-06,
+      "loss": 0.2021,
+      "step": 793
+    },
+    {
+      "epoch": 0.07727493917274939,
+      "grad_norm": 1.8762664332677217,
+      "learning_rate": 9.941677604841696e-06,
+      "loss": 0.6742,
+      "step": 794
+    },
+    {
+      "epoch": 0.07737226277372262,
+      "grad_norm": 1.5933514264940258,
+      "learning_rate": 9.9414373232494e-06,
+      "loss": 0.5156,
+      "step": 795
+    },
+    {
+      "epoch": 0.07746958637469586,
+      "grad_norm": 1.538651154827247,
+      "learning_rate": 9.94119655062344e-06,
+      "loss": 0.446,
+      "step": 796
+    },
+    {
+      "epoch": 0.0775669099756691,
+      "grad_norm": 3.7300878200470926,
+      "learning_rate": 9.94095528698774e-06,
+      "loss": 0.2745,
+      "step": 797
+    },
+    {
+      "epoch": 0.07766423357664233,
+      "grad_norm": 1.685774804326696,
+      "learning_rate": 9.940713532366277e-06,
+      "loss": 0.4236,
+      "step": 798
+    },
+    {
+      "epoch": 0.07776155717761557,
+      "grad_norm": 1.2528388212678458,
+      "learning_rate": 9.940471286783074e-06,
+      "loss": 0.308,
+      "step": 799
+    },
+    {
+      "epoch": 0.07785888077858881,
+      "grad_norm": 1.5082779398207746,
+      "learning_rate": 9.940228550262203e-06,
+      "loss": 0.4925,
+      "step": 800
+    },
+    {
+      "epoch": 0.07795620437956205,
+      "grad_norm": 1.544326069333433,
+      "learning_rate": 9.939985322827784e-06,
+      "loss": 0.4341,
+      "step": 801
+    },
+    {
+      "epoch": 0.07805352798053528,
+      "grad_norm": 1.4959220289677864,
+      "learning_rate": 9.939741604503987e-06,
+      "loss": 0.4548,
+      "step": 802
+    },
+    {
+      "epoch": 0.07815085158150852,
+      "grad_norm": 1.682287714178995,
+      "learning_rate": 9.93949739531503e-06,
+      "loss": 0.5277,
+      "step": 803
+    },
+    {
+      "epoch": 0.07824817518248176,
+      "grad_norm": 1.6519496438708445,
+      "learning_rate": 9.93925269528518e-06,
+      "loss": 0.3074,
+      "step": 804
+    },
+    {
+      "epoch": 0.07834549878345498,
+      "grad_norm": 1.4379883641500402,
+      "learning_rate": 9.939007504438756e-06,
+      "loss": 0.3069,
+      "step": 805
+    },
+    {
+      "epoch": 0.07844282238442822,
+      "grad_norm": 2.0644552037743793,
+      "learning_rate": 9.93876182280012e-06,
+      "loss": 0.4479,
+      "step": 806
+    },
+    {
+      "epoch": 0.07854014598540146,
+      "grad_norm": 1.4791313310441092,
+      "learning_rate": 9.938515650393685e-06,
+      "loss": 0.4255,
+      "step": 807
+    },
+    {
+      "epoch": 0.0786374695863747,
+      "grad_norm": 1.4280736600967436,
+      "learning_rate": 9.938268987243914e-06,
+      "loss": 0.466,
+      "step": 808
+    },
+    {
+      "epoch": 0.07873479318734793,
+      "grad_norm": 1.610976672135659,
+      "learning_rate": 9.93802183337532e-06,
+      "loss": 0.4327,
+      "step": 809
+    },
+    {
+      "epoch": 0.07883211678832117,
+      "grad_norm": 1.5447130604673693,
+      "learning_rate": 9.93777418881246e-06,
+      "loss": 0.4931,
+      "step": 810
+    },
+    {
+      "epoch": 0.0789294403892944,
+      "grad_norm": 1.3831325957946852,
+      "learning_rate": 9.937526053579944e-06,
+      "loss": 0.3877,
+      "step": 811
+    },
+    {
+      "epoch": 0.07902676399026765,
+      "grad_norm": 1.4247112282736865,
+      "learning_rate": 9.93727742770243e-06,
+      "loss": 0.4168,
+      "step": 812
+    },
+    {
+      "epoch": 0.07912408759124087,
+      "grad_norm": 1.5074130304911886,
+      "learning_rate": 9.937028311204624e-06,
+      "loss": 0.4747,
+      "step": 813
+    },
+    {
+      "epoch": 0.07922141119221411,
+      "grad_norm": 1.4955958242475926,
+      "learning_rate": 9.936778704111278e-06,
+      "loss": 0.2999,
+      "step": 814
+    },
+    {
+      "epoch": 0.07931873479318735,
+      "grad_norm": 1.6038468607718186,
+      "learning_rate": 9.9365286064472e-06,
+      "loss": 0.4897,
+      "step": 815
+    },
+    {
+      "epoch": 0.07941605839416059,
+      "grad_norm": 1.8040845780349017,
+      "learning_rate": 9.93627801823724e-06,
+      "loss": 0.6413,
+      "step": 816
+    },
+    {
+      "epoch": 0.07951338199513382,
+      "grad_norm": 1.4598215502284355,
+      "learning_rate": 9.936026939506298e-06,
+      "loss": 0.3687,
+      "step": 817
+    },
+    {
+      "epoch": 0.07961070559610706,
+      "grad_norm": 1.340412030499075,
+      "learning_rate": 9.935775370279324e-06,
+      "loss": 0.3833,
+      "step": 818
+    },
+    {
+      "epoch": 0.0797080291970803,
+      "grad_norm": 1.6913032059853774,
+      "learning_rate": 9.935523310581318e-06,
+      "loss": 0.5857,
+      "step": 819
+    },
+    {
+      "epoch": 0.07980535279805352,
+      "grad_norm": 1.9970663728185467,
+      "learning_rate": 9.93527076043733e-06,
+      "loss": 0.6843,
+      "step": 820
+    },
+    {
+      "epoch": 0.07990267639902676,
+      "grad_norm": 1.4408921562941295,
+      "learning_rate": 9.93501771987245e-06,
+      "loss": 0.4385,
+      "step": 821
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5184490203891443,
+      "learning_rate": 9.934764188911827e-06,
+      "loss": 0.4708,
+      "step": 822
+    },
+    {
+      "epoch": 0.08009732360097324,
+      "grad_norm": 1.8501562903086661,
+      "learning_rate": 9.934510167580654e-06,
+      "loss": 0.6431,
+      "step": 823
+    },
+    {
+      "epoch": 0.08019464720194647,
+      "grad_norm": 1.6997829158405129,
+      "learning_rate": 9.934255655904172e-06,
+      "loss": 0.5188,
+      "step": 824
+    },
+    {
+      "epoch": 0.08029197080291971,
+      "grad_norm": 1.8510241792275326,
+      "learning_rate": 9.934000653907674e-06,
+      "loss": 0.5457,
+      "step": 825
+    },
+    {
+      "epoch": 0.08038929440389295,
+      "grad_norm": 1.6853569692908912,
+      "learning_rate": 9.933745161616498e-06,
+      "loss": 0.5062,
+      "step": 826
+    },
+    {
+      "epoch": 0.08048661800486619,
+      "grad_norm": 1.3066104263898661,
+      "learning_rate": 9.93348917905603e-06,
+      "loss": 0.404,
+      "step": 827
+    },
+    {
+      "epoch": 0.08058394160583941,
+      "grad_norm": 1.2788244408859646,
+      "learning_rate": 9.933232706251712e-06,
+      "loss": 0.3253,
+      "step": 828
+    },
+    {
+      "epoch": 0.08068126520681265,
+      "grad_norm": 2.2690800072126325,
+      "learning_rate": 9.932975743229027e-06,
+      "loss": 0.3405,
+      "step": 829
+    },
+    {
+      "epoch": 0.08077858880778589,
+      "grad_norm": 1.9113871035353245,
+      "learning_rate": 9.932718290013512e-06,
+      "loss": 0.5989,
+      "step": 830
+    },
+    {
+      "epoch": 0.08087591240875912,
+      "grad_norm": 1.3655256798283997,
+      "learning_rate": 9.932460346630748e-06,
+      "loss": 0.2942,
+      "step": 831
+    },
+    {
+      "epoch": 0.08097323600973236,
+      "grad_norm": 1.5234864838378999,
+      "learning_rate": 9.932201913106366e-06,
+      "loss": 0.3913,
+      "step": 832
+    },
+    {
+      "epoch": 0.0810705596107056,
+      "grad_norm": 1.3752195876516826,
+      "learning_rate": 9.93194298946605e-06,
+      "loss": 0.3293,
+      "step": 833
+    },
+    {
+      "epoch": 0.08116788321167884,
+      "grad_norm": 1.4842622412969824,
+      "learning_rate": 9.931683575735527e-06,
+      "loss": 0.4157,
+      "step": 834
+    },
+    {
+      "epoch": 0.08126520681265206,
+      "grad_norm": 4.003685207313109,
+      "learning_rate": 9.931423671940577e-06,
+      "loss": 0.3276,
+      "step": 835
+    },
+    {
+      "epoch": 0.0813625304136253,
+      "grad_norm": 1.509943035011216,
+      "learning_rate": 9.931163278107023e-06,
+      "loss": 0.4045,
+      "step": 836
+    },
+    {
+      "epoch": 0.08145985401459854,
+      "grad_norm": 1.4382523765338775,
+      "learning_rate": 9.930902394260746e-06,
+      "loss": 0.2709,
+      "step": 837
+    },
+    {
+      "epoch": 0.08155717761557178,
+      "grad_norm": 1.4492711471586157,
+      "learning_rate": 9.930641020427665e-06,
+      "loss": 0.3957,
+      "step": 838
+    },
+    {
+      "epoch": 0.08165450121654501,
+      "grad_norm": 1.7428876214187694,
+      "learning_rate": 9.930379156633758e-06,
+      "loss": 0.5257,
+      "step": 839
+    },
+    {
+      "epoch": 0.08175182481751825,
+      "grad_norm": 1.5652514836380926,
+      "learning_rate": 9.930116802905042e-06,
+      "loss": 0.4948,
+      "step": 840
+    },
+    {
+      "epoch": 0.08184914841849149,
+      "grad_norm": 2.4133112951540494,
+      "learning_rate": 9.929853959267589e-06,
+      "loss": 0.5455,
+      "step": 841
+    },
+    {
+      "epoch": 0.08194647201946471,
+      "grad_norm": 1.4309460046419233,
+      "learning_rate": 9.929590625747518e-06,
+      "loss": 0.4057,
+      "step": 842
+    },
+    {
+      "epoch": 0.08204379562043795,
+      "grad_norm": 1.0450296792009146,
+      "learning_rate": 9.929326802370995e-06,
+      "loss": 0.2332,
+      "step": 843
+    },
+    {
+      "epoch": 0.0821411192214112,
+      "grad_norm": 1.1201933325217828,
+      "learning_rate": 9.92906248916424e-06,
+      "loss": 0.3264,
+      "step": 844
+    },
+    {
+      "epoch": 0.08223844282238443,
+      "grad_norm": 1.6243579769967154,
+      "learning_rate": 9.928797686153515e-06,
+      "loss": 0.5385,
+      "step": 845
+    },
+    {
+      "epoch": 0.08233576642335766,
+      "grad_norm": 1.3496069901220336,
+      "learning_rate": 9.928532393365136e-06,
+      "loss": 0.3875,
+      "step": 846
+    },
+    {
+      "epoch": 0.0824330900243309,
+      "grad_norm": 1.4862888245769246,
+      "learning_rate": 9.928266610825462e-06,
+      "loss": 0.4493,
+      "step": 847
+    },
+    {
+      "epoch": 0.08253041362530414,
+      "grad_norm": 1.8305160014899666,
+      "learning_rate": 9.928000338560906e-06,
+      "loss": 0.4582,
+      "step": 848
+    },
+    {
+      "epoch": 0.08262773722627738,
+      "grad_norm": 1.642584946989029,
+      "learning_rate": 9.927733576597926e-06,
+      "loss": 0.3347,
+      "step": 849
+    },
+    {
+      "epoch": 0.0827250608272506,
+      "grad_norm": 1.5413363162928122,
+      "learning_rate": 9.927466324963033e-06,
+      "loss": 0.4607,
+      "step": 850
+    },
+    {
+      "epoch": 0.08282238442822384,
+      "grad_norm": 1.7093263469236866,
+      "learning_rate": 9.927198583682784e-06,
+      "loss": 0.5706,
+      "step": 851
+    },
+    {
+      "epoch": 0.08291970802919708,
+      "grad_norm": 1.531714933227777,
+      "learning_rate": 9.926930352783781e-06,
+      "loss": 0.533,
+      "step": 852
+    },
+    {
+      "epoch": 0.08301703163017031,
+      "grad_norm": 1.8181822267445191,
+      "learning_rate": 9.926661632292683e-06,
+      "loss": 0.5946,
+      "step": 853
+    },
+    {
+      "epoch": 0.08311435523114355,
+      "grad_norm": 1.8304662465930317,
+      "learning_rate": 9.926392422236189e-06,
+      "loss": 0.3746,
+      "step": 854
+    },
+    {
+      "epoch": 0.08321167883211679,
+      "grad_norm": 1.3135536142885351,
+      "learning_rate": 9.926122722641051e-06,
+      "loss": 0.429,
+      "step": 855
+    },
+    {
+      "epoch": 0.08330900243309003,
+      "grad_norm": 1.714390027755308,
+      "learning_rate": 9.925852533534071e-06,
+      "loss": 0.6806,
+      "step": 856
+    },
+    {
+      "epoch": 0.08340632603406326,
+      "grad_norm": 1.3399957064659453,
+      "learning_rate": 9.925581854942099e-06,
+      "loss": 0.2824,
+      "step": 857
+    },
+    {
+      "epoch": 0.0835036496350365,
+      "grad_norm": 1.3705351036499993,
+      "learning_rate": 9.925310686892026e-06,
+      "loss": 0.3085,
+      "step": 858
+    },
+    {
+      "epoch": 0.08360097323600973,
+      "grad_norm": 1.5064665959171673,
+      "learning_rate": 9.925039029410807e-06,
+      "loss": 0.4445,
+      "step": 859
+    },
+    {
+      "epoch": 0.08369829683698297,
+      "grad_norm": 1.725614330530946,
+      "learning_rate": 9.924766882525433e-06,
+      "loss": 0.4704,
+      "step": 860
+    },
+    {
+      "epoch": 0.0837956204379562,
+      "grad_norm": 1.765372064078189,
+      "learning_rate": 9.924494246262944e-06,
+      "loss": 0.6383,
+      "step": 861
+    },
+    {
+      "epoch": 0.08389294403892944,
+      "grad_norm": 2.085503007877936,
+      "learning_rate": 9.924221120650434e-06,
+      "loss": 0.296,
+      "step": 862
+    },
+    {
+      "epoch": 0.08399026763990268,
+      "grad_norm": 1.7898541160892734,
+      "learning_rate": 9.923947505715046e-06,
+      "loss": 0.5991,
+      "step": 863
+    },
+    {
+      "epoch": 0.0840875912408759,
+      "grad_norm": 1.6476104975968628,
+      "learning_rate": 9.923673401483968e-06,
+      "loss": 0.4734,
+      "step": 864
+    },
+    {
+      "epoch": 0.08418491484184915,
+      "grad_norm": 1.5502768976775265,
+      "learning_rate": 9.923398807984439e-06,
+      "loss": 0.2764,
+      "step": 865
+    },
+    {
+      "epoch": 0.08428223844282239,
+      "grad_norm": 1.2398437846135097,
+      "learning_rate": 9.923123725243744e-06,
+      "loss": 0.2705,
+      "step": 866
+    },
+    {
+      "epoch": 0.08437956204379563,
+      "grad_norm": 1.5290591078236662,
+      "learning_rate": 9.922848153289217e-06,
+      "loss": 0.4228,
+      "step": 867
+    },
+    {
+      "epoch": 0.08447688564476885,
+      "grad_norm": 1.134889947118225,
+      "learning_rate": 9.922572092148244e-06,
+      "loss": 0.2953,
+      "step": 868
+    },
+    {
+      "epoch": 0.08457420924574209,
+      "grad_norm": 1.6307620082274505,
+      "learning_rate": 9.922295541848257e-06,
+      "loss": 0.3363,
+      "step": 869
+    },
+    {
+      "epoch": 0.08467153284671533,
+      "grad_norm": 1.373015271795792,
+      "learning_rate": 9.922018502416736e-06,
+      "loss": 0.3593,
+      "step": 870
+    },
+    {
+      "epoch": 0.08476885644768857,
+      "grad_norm": 1.7500724096304088,
+      "learning_rate": 9.921740973881211e-06,
+      "loss": 0.5236,
+      "step": 871
+    },
+    {
+      "epoch": 0.0848661800486618,
+      "grad_norm": 1.6167507595463353,
+      "learning_rate": 9.92146295626926e-06,
+      "loss": 0.5138,
+      "step": 872
+    },
+    {
+      "epoch": 0.08496350364963504,
+      "grad_norm": 1.0398007401901226,
+      "learning_rate": 9.92118444960851e-06,
+      "loss": 0.295,
+      "step": 873
+    },
+    {
+      "epoch": 0.08506082725060828,
+      "grad_norm": 1.4140920056378707,
+      "learning_rate": 9.920905453926637e-06,
+      "loss": 0.4192,
+      "step": 874
+    },
+    {
+      "epoch": 0.0851581508515815,
+      "grad_norm": 1.8785238213855096,
+      "learning_rate": 9.920625969251365e-06,
+      "loss": 0.4228,
+      "step": 875
+    },
+    {
+      "epoch": 0.08525547445255474,
+      "grad_norm": 1.719991686268608,
+      "learning_rate": 9.920345995610465e-06,
+      "loss": 0.5026,
+      "step": 876
+    },
+    {
+      "epoch": 0.08535279805352798,
+      "grad_norm": 1.7112372148926476,
+      "learning_rate": 9.92006553303176e-06,
+      "loss": 0.3157,
+      "step": 877
+    },
+    {
+      "epoch": 0.08545012165450122,
+      "grad_norm": 2.5105720144829116,
+      "learning_rate": 9.919784581543117e-06,
+      "loss": 0.4777,
+      "step": 878
+    },
+    {
+      "epoch": 0.08554744525547445,
+      "grad_norm": 1.42848630379055,
+      "learning_rate": 9.919503141172458e-06,
+      "loss": 0.3998,
+      "step": 879
+    },
+    {
+      "epoch": 0.08564476885644769,
+      "grad_norm": 1.4246136626839867,
+      "learning_rate": 9.919221211947748e-06,
+      "loss": 0.4415,
+      "step": 880
+    },
+    {
+      "epoch": 0.08574209245742093,
+      "grad_norm": 1.939970471855472,
+      "learning_rate": 9.918938793897002e-06,
+      "loss": 0.5887,
+      "step": 881
+    },
+    {
+      "epoch": 0.08583941605839417,
+      "grad_norm": 1.5467402852284964,
+      "learning_rate": 9.918655887048285e-06,
+      "loss": 0.3726,
+      "step": 882
+    },
+    {
+      "epoch": 0.08593673965936739,
+      "grad_norm": 1.6261636529000345,
+      "learning_rate": 9.918372491429708e-06,
+      "loss": 0.3382,
+      "step": 883
+    },
+    {
+      "epoch": 0.08603406326034063,
+      "grad_norm": 1.4859289768748727,
+      "learning_rate": 9.918088607069434e-06,
+      "loss": 0.4837,
+      "step": 884
+    },
+    {
+      "epoch": 0.08613138686131387,
+      "grad_norm": 1.8534453271170916,
+      "learning_rate": 9.917804233995673e-06,
+      "loss": 0.5948,
+      "step": 885
+    },
+    {
+      "epoch": 0.08622871046228711,
+      "grad_norm": 1.3491809126204122,
+      "learning_rate": 9.917519372236684e-06,
+      "loss": 0.381,
+      "step": 886
+    },
+    {
+      "epoch": 0.08632603406326034,
+      "grad_norm": 1.4913268478302555,
+      "learning_rate": 9.91723402182077e-06,
+      "loss": 0.2872,
+      "step": 887
+    },
+    {
+      "epoch": 0.08642335766423358,
+      "grad_norm": 1.5345667515291348,
+      "learning_rate": 9.916948182776289e-06,
+      "loss": 0.4426,
+      "step": 888
+    },
+    {
+      "epoch": 0.08652068126520682,
+      "grad_norm": 1.9142340135608018,
+      "learning_rate": 9.916661855131646e-06,
+      "loss": 0.467,
+      "step": 889
+    },
+    {
+      "epoch": 0.08661800486618004,
+      "grad_norm": 1.7451883652681546,
+      "learning_rate": 9.916375038915291e-06,
+      "loss": 0.3579,
+      "step": 890
+    },
+    {
+      "epoch": 0.08671532846715328,
+      "grad_norm": 3.3675828599824618,
+      "learning_rate": 9.916087734155728e-06,
+      "loss": 0.3965,
+      "step": 891
+    },
+    {
+      "epoch": 0.08681265206812652,
+      "grad_norm": 1.6430989821947144,
+      "learning_rate": 9.915799940881504e-06,
+      "loss": 0.5089,
+      "step": 892
+    },
+    {
+      "epoch": 0.08690997566909976,
+      "grad_norm": 1.8434153107573372,
+      "learning_rate": 9.915511659121219e-06,
+      "loss": 0.6513,
+      "step": 893
+    },
+    {
+      "epoch": 0.08700729927007299,
+      "grad_norm": 1.7259560464984558,
+      "learning_rate": 9.91522288890352e-06,
+      "loss": 0.5963,
+      "step": 894
+    },
+    {
+      "epoch": 0.08710462287104623,
+      "grad_norm": 1.4417036209809253,
+      "learning_rate": 9.9149336302571e-06,
+      "loss": 0.4076,
+      "step": 895
+    },
+    {
+      "epoch": 0.08720194647201947,
+      "grad_norm": 1.4565626930182671,
+      "learning_rate": 9.914643883210704e-06,
+      "loss": 0.3548,
+      "step": 896
+    },
+    {
+      "epoch": 0.08729927007299271,
+      "grad_norm": 1.8286482885292266,
+      "learning_rate": 9.914353647793126e-06,
+      "loss": 0.5158,
+      "step": 897
+    },
+    {
+      "epoch": 0.08739659367396593,
+      "grad_norm": 1.573235746781315,
+      "learning_rate": 9.914062924033204e-06,
+      "loss": 0.4804,
+      "step": 898
+    },
+    {
+      "epoch": 0.08749391727493917,
+      "grad_norm": 1.7725042500734154,
+      "learning_rate": 9.91377171195983e-06,
+      "loss": 0.4037,
+      "step": 899
+    },
+    {
+      "epoch": 0.08759124087591241,
+      "grad_norm": 1.5572801757524644,
+      "learning_rate": 9.913480011601939e-06,
+      "loss": 0.2757,
+      "step": 900
+    },
+    {
+      "epoch": 0.08768856447688564,
+      "grad_norm": 1.690990088453521,
+      "learning_rate": 9.91318782298852e-06,
+      "loss": 0.624,
+      "step": 901
+    },
+    {
+      "epoch": 0.08778588807785888,
+      "grad_norm": 1.5797017595834213,
+      "learning_rate": 9.912895146148609e-06,
+      "loss": 0.418,
+      "step": 902
+    },
+    {
+      "epoch": 0.08788321167883212,
+      "grad_norm": 1.722754374021215,
+      "learning_rate": 9.912601981111287e-06,
+      "loss": 0.5991,
+      "step": 903
+    },
+    {
+      "epoch": 0.08798053527980536,
+      "grad_norm": 1.2395740583484196,
+      "learning_rate": 9.912308327905683e-06,
+      "loss": 0.3632,
+      "step": 904
+    },
+    {
+      "epoch": 0.08807785888077858,
+      "grad_norm": 1.8637568028899596,
+      "learning_rate": 9.912014186560985e-06,
+      "loss": 0.5766,
+      "step": 905
+    },
+    {
+      "epoch": 0.08817518248175182,
+      "grad_norm": 1.8489319991981024,
+      "learning_rate": 9.911719557106418e-06,
+      "loss": 0.6834,
+      "step": 906
+    },
+    {
+      "epoch": 0.08827250608272506,
+      "grad_norm": 1.6692858460733677,
+      "learning_rate": 9.911424439571258e-06,
+      "loss": 0.5067,
+      "step": 907
+    },
+    {
+      "epoch": 0.0883698296836983,
+      "grad_norm": 1.4727605888984552,
+      "learning_rate": 9.911128833984834e-06,
+      "loss": 0.3141,
+      "step": 908
+    },
+    {
+      "epoch": 0.08846715328467153,
+      "grad_norm": 1.644393806422472,
+      "learning_rate": 9.910832740376518e-06,
+      "loss": 0.4599,
+      "step": 909
+    },
+    {
+      "epoch": 0.08856447688564477,
+      "grad_norm": 1.730275300452632,
+      "learning_rate": 9.910536158775734e-06,
+      "loss": 0.3908,
+      "step": 910
+    },
+    {
+      "epoch": 0.08866180048661801,
+      "grad_norm": 1.7281903494262714,
+      "learning_rate": 9.910239089211955e-06,
+      "loss": 0.5919,
+      "step": 911
+    },
+    {
+      "epoch": 0.08875912408759123,
+      "grad_norm": 1.7234172913238917,
+      "learning_rate": 9.909941531714699e-06,
+      "loss": 0.609,
+      "step": 912
+    },
+    {
+      "epoch": 0.08885644768856447,
+      "grad_norm": 1.4594702058569258,
+      "learning_rate": 9.909643486313533e-06,
+      "loss": 0.4399,
+      "step": 913
+    },
+    {
+      "epoch": 0.08895377128953771,
+      "grad_norm": 1.4625782448468165,
+      "learning_rate": 9.90934495303808e-06,
+      "loss": 0.4011,
+      "step": 914
+    },
+    {
+      "epoch": 0.08905109489051095,
+      "grad_norm": 1.7262645481609784,
+      "learning_rate": 9.909045931918e-06,
+      "loss": 0.4992,
+      "step": 915
+    },
+    {
+      "epoch": 0.08914841849148418,
+      "grad_norm": 1.6255222361700263,
+      "learning_rate": 9.908746422983007e-06,
+      "loss": 0.4909,
+      "step": 916
+    },
+    {
+      "epoch": 0.08924574209245742,
+      "grad_norm": 1.7512982185254946,
+      "learning_rate": 9.908446426262865e-06,
+      "loss": 0.5527,
+      "step": 917
+    },
+    {
+      "epoch": 0.08934306569343066,
+      "grad_norm": 1.617605772613541,
+      "learning_rate": 9.908145941787386e-06,
+      "loss": 0.3228,
+      "step": 918
+    },
+    {
+      "epoch": 0.0894403892944039,
+      "grad_norm": 1.489706963519404,
+      "learning_rate": 9.907844969586427e-06,
+      "loss": 0.4838,
+      "step": 919
+    },
+    {
+      "epoch": 0.08953771289537713,
+      "grad_norm": 1.193837371345013,
+      "learning_rate": 9.907543509689896e-06,
+      "loss": 0.284,
+      "step": 920
+    },
+    {
+      "epoch": 0.08963503649635036,
+      "grad_norm": 1.5855787651349198,
+      "learning_rate": 9.907241562127752e-06,
+      "loss": 0.4641,
+      "step": 921
+    },
+    {
+      "epoch": 0.0897323600973236,
+      "grad_norm": 1.2401284480478103,
+      "learning_rate": 9.906939126929998e-06,
+      "loss": 0.246,
+      "step": 922
+    },
+    {
+      "epoch": 0.08982968369829683,
+      "grad_norm": 1.503842201355298,
+      "learning_rate": 9.906636204126685e-06,
+      "loss": 0.4031,
+      "step": 923
+    },
+    {
+      "epoch": 0.08992700729927007,
+      "grad_norm": 1.9138265658958267,
+      "learning_rate": 9.906332793747917e-06,
+      "loss": 0.587,
+      "step": 924
+    },
+    {
+      "epoch": 0.09002433090024331,
+      "grad_norm": 1.5381184892388742,
+      "learning_rate": 9.906028895823844e-06,
+      "loss": 0.4119,
+      "step": 925
+    },
+    {
+      "epoch": 0.09012165450121655,
+      "grad_norm": 1.5769181877690257,
+      "learning_rate": 9.905724510384664e-06,
+      "loss": 0.4071,
+      "step": 926
+    },
+    {
+      "epoch": 0.09021897810218978,
+      "grad_norm": 1.4644408625641083,
+      "learning_rate": 9.905419637460625e-06,
+      "loss": 0.3656,
+      "step": 927
+    },
+    {
+      "epoch": 0.09031630170316302,
+      "grad_norm": 2.043739071504731,
+      "learning_rate": 9.90511427708202e-06,
+      "loss": 0.6317,
+      "step": 928
+    },
+    {
+      "epoch": 0.09041362530413626,
+      "grad_norm": 1.8397228419915481,
+      "learning_rate": 9.904808429279195e-06,
+      "loss": 0.6656,
+      "step": 929
+    },
+    {
+      "epoch": 0.0905109489051095,
+      "grad_norm": 1.6689588837493128,
+      "learning_rate": 9.904502094082542e-06,
+      "loss": 0.4603,
+      "step": 930
+    },
+    {
+      "epoch": 0.09060827250608272,
+      "grad_norm": 1.7157610479724803,
+      "learning_rate": 9.9041952715225e-06,
+      "loss": 0.3566,
+      "step": 931
+    },
+    {
+      "epoch": 0.09070559610705596,
+      "grad_norm": 1.5797548847560638,
+      "learning_rate": 9.90388796162956e-06,
+      "loss": 0.527,
+      "step": 932
+    },
+    {
+      "epoch": 0.0908029197080292,
+      "grad_norm": 1.3861944362556795,
+      "learning_rate": 9.903580164434262e-06,
+      "loss": 0.3555,
+      "step": 933
+    },
+    {
+      "epoch": 0.09090024330900243,
+      "grad_norm": 1.4873043668950738,
+      "learning_rate": 9.903271879967185e-06,
+      "loss": 0.3606,
+      "step": 934
+    },
+    {
+      "epoch": 0.09099756690997567,
+      "grad_norm": 1.5471770637050817,
+      "learning_rate": 9.90296310825897e-06,
+      "loss": 0.5407,
+      "step": 935
+    },
+    {
+      "epoch": 0.0910948905109489,
+      "grad_norm": 1.7410898214633266,
+      "learning_rate": 9.902653849340296e-06,
+      "loss": 0.5604,
+      "step": 936
+    },
+    {
+      "epoch": 0.09119221411192215,
+      "grad_norm": 1.490257412993615,
+      "learning_rate": 9.902344103241897e-06,
+      "loss": 0.4293,
+      "step": 937
+    },
+    {
+      "epoch": 0.09128953771289537,
+      "grad_norm": 1.3076716120407041,
+      "learning_rate": 9.90203386999455e-06,
+      "loss": 0.4311,
+      "step": 938
+    },
+    {
+      "epoch": 0.09138686131386861,
+      "grad_norm": 1.63883307554104,
+      "learning_rate": 9.901723149629085e-06,
+      "loss": 0.5026,
+      "step": 939
+    },
+    {
+      "epoch": 0.09148418491484185,
+      "grad_norm": 1.460694807977355,
+      "learning_rate": 9.901411942176377e-06,
+      "loss": 0.4449,
+      "step": 940
+    },
+    {
+      "epoch": 0.09158150851581509,
+      "grad_norm": 1.631318499416747,
+      "learning_rate": 9.901100247667352e-06,
+      "loss": 0.4762,
+      "step": 941
+    },
+    {
+      "epoch": 0.09167883211678832,
+      "grad_norm": 1.472942456024595,
+      "learning_rate": 9.900788066132982e-06,
+      "loss": 0.4208,
+      "step": 942
+    },
+    {
+      "epoch": 0.09177615571776156,
+      "grad_norm": 1.9471723252943203,
+      "learning_rate": 9.900475397604292e-06,
+      "loss": 0.4887,
+      "step": 943
+    },
+    {
+      "epoch": 0.0918734793187348,
+      "grad_norm": 1.4192635165617975,
+      "learning_rate": 9.900162242112348e-06,
+      "loss": 0.4753,
+      "step": 944
+    },
+    {
+      "epoch": 0.09197080291970802,
+      "grad_norm": 1.7864248496903834,
+      "learning_rate": 9.89984859968827e-06,
+      "loss": 0.6063,
+      "step": 945
+    },
+    {
+      "epoch": 0.09206812652068126,
+      "grad_norm": 1.402919088092856,
+      "learning_rate": 9.899534470363225e-06,
+      "loss": 0.3561,
+      "step": 946
+    },
+    {
+      "epoch": 0.0921654501216545,
+      "grad_norm": 1.15011785152118,
+      "learning_rate": 9.89921985416843e-06,
+      "loss": 0.2605,
+      "step": 947
+    },
+    {
+      "epoch": 0.09226277372262774,
+      "grad_norm": 1.2940536511249239,
+      "learning_rate": 9.898904751135145e-06,
+      "loss": 0.2503,
+      "step": 948
+    },
+    {
+      "epoch": 0.09236009732360097,
+      "grad_norm": 1.5093308152075566,
+      "learning_rate": 9.898589161294684e-06,
+      "loss": 0.4185,
+      "step": 949
+    },
+    {
+      "epoch": 0.09245742092457421,
+      "grad_norm": 1.5826010349075055,
+      "learning_rate": 9.898273084678406e-06,
+      "loss": 0.536,
+      "step": 950
+    },
+    {
+      "epoch": 0.09255474452554745,
+      "grad_norm": 1.5672518381317015,
+      "learning_rate": 9.897956521317724e-06,
+      "loss": 0.5068,
+      "step": 951
+    },
+    {
+      "epoch": 0.09265206812652069,
+      "grad_norm": 1.784767292144658,
+      "learning_rate": 9.89763947124409e-06,
+      "loss": 0.6601,
+      "step": 952
+    },
+    {
+      "epoch": 0.09274939172749391,
+      "grad_norm": 1.620681747107968,
+      "learning_rate": 9.897321934489011e-06,
+      "loss": 0.5402,
+      "step": 953
+    },
+    {
+      "epoch": 0.09284671532846715,
+      "grad_norm": 1.7479722673062432,
+      "learning_rate": 9.897003911084042e-06,
+      "loss": 0.6593,
+      "step": 954
+    },
+    {
+      "epoch": 0.09294403892944039,
+      "grad_norm": 1.6618363798373263,
+      "learning_rate": 9.896685401060783e-06,
+      "loss": 0.6086,
+      "step": 955
+    },
+    {
+      "epoch": 0.09304136253041363,
+      "grad_norm": 1.3782603882872615,
+      "learning_rate": 9.896366404450888e-06,
+      "loss": 0.3431,
+      "step": 956
+    },
+    {
+      "epoch": 0.09313868613138686,
+      "grad_norm": 1.6607836446620106,
+      "learning_rate": 9.896046921286053e-06,
+      "loss": 0.4015,
+      "step": 957
+    },
+    {
+      "epoch": 0.0932360097323601,
+      "grad_norm": 1.372535143543006,
+      "learning_rate": 9.895726951598026e-06,
+      "loss": 0.3627,
+      "step": 958
+    },
+    {
+      "epoch": 0.09333333333333334,
+      "grad_norm": 1.965175835699204,
+      "learning_rate": 9.895406495418602e-06,
+      "loss": 0.434,
+      "step": 959
+    },
+    {
+      "epoch": 0.09343065693430656,
+      "grad_norm": 1.6072227382486934,
+      "learning_rate": 9.895085552779626e-06,
+      "loss": 0.3666,
+      "step": 960
+    },
+    {
+      "epoch": 0.0935279805352798,
+      "grad_norm": 1.8680414138630521,
+      "learning_rate": 9.894764123712991e-06,
+      "loss": 0.6182,
+      "step": 961
+    },
+    {
+      "epoch": 0.09362530413625304,
+      "grad_norm": 1.7249394724081422,
+      "learning_rate": 9.894442208250636e-06,
+      "loss": 0.569,
+      "step": 962
+    },
+    {
+      "epoch": 0.09372262773722628,
+      "grad_norm": 1.7887658285510963,
+      "learning_rate": 9.894119806424549e-06,
+      "loss": 0.4825,
+      "step": 963
+    },
+    {
+      "epoch": 0.09381995133819951,
+      "grad_norm": 1.4470695743772581,
+      "learning_rate": 9.89379691826677e-06,
+      "loss": 0.4036,
+      "step": 964
+    },
+    {
+      "epoch": 0.09391727493917275,
+      "grad_norm": 1.739037372856574,
+      "learning_rate": 9.893473543809383e-06,
+      "loss": 0.3734,
+      "step": 965
+    },
+    {
+      "epoch": 0.09401459854014599,
+      "grad_norm": 1.2401623802615098,
+      "learning_rate": 9.893149683084522e-06,
+      "loss": 0.2892,
+      "step": 966
+    },
+    {
+      "epoch": 0.09411192214111923,
+      "grad_norm": 1.632367817316159,
+      "learning_rate": 9.892825336124369e-06,
+      "loss": 0.3324,
+      "step": 967
+    },
+    {
+      "epoch": 0.09420924574209245,
+      "grad_norm": 1.4553279790204596,
+      "learning_rate": 9.892500502961156e-06,
+      "loss": 0.4518,
+      "step": 968
+    },
+    {
+      "epoch": 0.0943065693430657,
+      "grad_norm": 2.0184949211791867,
+      "learning_rate": 9.892175183627161e-06,
+      "loss": 0.496,
+      "step": 969
+    },
+    {
+      "epoch": 0.09440389294403893,
+      "grad_norm": 1.3847811204395728,
+      "learning_rate": 9.89184937815471e-06,
+      "loss": 0.3908,
+      "step": 970
+    },
+    {
+      "epoch": 0.09450121654501216,
+      "grad_norm": 1.7325451795183482,
+      "learning_rate": 9.89152308657618e-06,
+      "loss": 0.5813,
+      "step": 971
+    },
+    {
+      "epoch": 0.0945985401459854,
+      "grad_norm": 1.3485480854398895,
+      "learning_rate": 9.891196308923994e-06,
+      "loss": 0.2773,
+      "step": 972
+    },
+    {
+      "epoch": 0.09469586374695864,
+      "grad_norm": 1.6137214411092917,
+      "learning_rate": 9.890869045230625e-06,
+      "loss": 0.573,
+      "step": 973
+    },
+    {
+      "epoch": 0.09479318734793188,
+      "grad_norm": 1.8098732560393935,
+      "learning_rate": 9.890541295528593e-06,
+      "loss": 0.5765,
+      "step": 974
+    },
+    {
+      "epoch": 0.0948905109489051,
+      "grad_norm": 1.7169741386061155,
+      "learning_rate": 9.890213059850467e-06,
+      "loss": 0.5463,
+      "step": 975
+    },
+    {
+      "epoch": 0.09498783454987834,
+      "grad_norm": 1.6226425677233698,
+      "learning_rate": 9.889884338228861e-06,
+      "loss": 0.459,
+      "step": 976
+    },
+    {
+      "epoch": 0.09508515815085158,
+      "grad_norm": 1.5712338302132318,
+      "learning_rate": 9.889555130696445e-06,
+      "loss": 0.2926,
+      "step": 977
+    },
+    {
+      "epoch": 0.09518248175182482,
+      "grad_norm": 2.368668096329164,
+      "learning_rate": 9.88922543728593e-06,
+      "loss": 0.4602,
+      "step": 978
+    },
+    {
+      "epoch": 0.09527980535279805,
+      "grad_norm": 1.5481463515619227,
+      "learning_rate": 9.888895258030077e-06,
+      "loss": 0.382,
+      "step": 979
+    },
+    {
+      "epoch": 0.09537712895377129,
+      "grad_norm": 1.5566394762827083,
+      "learning_rate": 9.888564592961698e-06,
+      "loss": 0.4432,
+      "step": 980
+    },
+    {
+      "epoch": 0.09547445255474453,
+      "grad_norm": 1.2929219586068095,
+      "learning_rate": 9.888233442113651e-06,
+      "loss": 0.2986,
+      "step": 981
+    },
+    {
+      "epoch": 0.09557177615571776,
+      "grad_norm": 1.7926346211976876,
+      "learning_rate": 9.887901805518841e-06,
+      "loss": 0.4536,
+      "step": 982
+    },
+    {
+      "epoch": 0.095669099756691,
+      "grad_norm": 1.5810862037952855,
+      "learning_rate": 9.887569683210225e-06,
+      "loss": 0.5143,
+      "step": 983
+    },
+    {
+      "epoch": 0.09576642335766423,
+      "grad_norm": 1.486412737689962,
+      "learning_rate": 9.887237075220805e-06,
+      "loss": 0.4422,
+      "step": 984
+    },
+    {
+      "epoch": 0.09586374695863747,
+      "grad_norm": 1.5634292890846626,
+      "learning_rate": 9.886903981583633e-06,
+      "loss": 0.5158,
+      "step": 985
+    },
+    {
+      "epoch": 0.0959610705596107,
+      "grad_norm": 1.4911106877832496,
+      "learning_rate": 9.88657040233181e-06,
+      "loss": 0.3584,
+      "step": 986
+    },
+    {
+      "epoch": 0.09605839416058394,
+      "grad_norm": 1.8920202230134835,
+      "learning_rate": 9.886236337498481e-06,
+      "loss": 0.7059,
+      "step": 987
+    },
+    {
+      "epoch": 0.09615571776155718,
+      "grad_norm": 1.9765830057761664,
+      "learning_rate": 9.885901787116844e-06,
+      "loss": 0.3363,
+      "step": 988
+    },
+    {
+      "epoch": 0.09625304136253042,
+      "grad_norm": 1.7412713212065478,
+      "learning_rate": 9.885566751220144e-06,
+      "loss": 0.6238,
+      "step": 989
+    },
+    {
+      "epoch": 0.09635036496350365,
+      "grad_norm": 1.4558500764026314,
+      "learning_rate": 9.885231229841675e-06,
+      "loss": 0.5033,
+      "step": 990
+    },
+    {
+      "epoch": 0.09644768856447689,
+      "grad_norm": 1.5722863237428275,
+      "learning_rate": 9.884895223014772e-06,
+      "loss": 0.3026,
+      "step": 991
+    },
+    {
+      "epoch": 0.09654501216545013,
+      "grad_norm": 1.7850396516814273,
+      "learning_rate": 9.88455873077283e-06,
+      "loss": 0.6797,
+      "step": 992
+    },
+    {
+      "epoch": 0.09664233576642335,
+      "grad_norm": 1.5907642595826164,
+      "learning_rate": 9.884221753149286e-06,
+      "loss": 0.5051,
+      "step": 993
+    },
+    {
+      "epoch": 0.09673965936739659,
+      "grad_norm": 1.383326117178851,
+      "learning_rate": 9.883884290177623e-06,
+      "loss": 0.394,
+      "step": 994
+    },
+    {
+      "epoch": 0.09683698296836983,
+      "grad_norm": 1.5330791836349085,
+      "learning_rate": 9.883546341891375e-06,
+      "loss": 0.4531,
+      "step": 995
+    },
+    {
+      "epoch": 0.09693430656934307,
+      "grad_norm": 1.3858453283442664,
+      "learning_rate": 9.883207908324126e-06,
+      "loss": 0.4674,
+      "step": 996
+    },
+    {
+      "epoch": 0.0970316301703163,
+      "grad_norm": 1.2633519423598012,
+      "learning_rate": 9.882868989509507e-06,
+      "loss": 0.3053,
+      "step": 997
+    },
+    {
+      "epoch": 0.09712895377128954,
+      "grad_norm": 1.5725755469000553,
+      "learning_rate": 9.882529585481194e-06,
+      "loss": 0.5382,
+      "step": 998
+    },
+    {
+      "epoch": 0.09722627737226278,
+      "grad_norm": 1.594807816051373,
+      "learning_rate": 9.882189696272916e-06,
+      "loss": 0.5027,
+      "step": 999
+    },
+    {
+      "epoch": 0.09732360097323602,
+      "grad_norm": 1.7855937930735857,
+      "learning_rate": 9.881849321918446e-06,
+      "loss": 0.6336,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09742092457420924,
+      "grad_norm": 1.8161736452208326,
+      "learning_rate": 9.88150846245161e-06,
+      "loss": 0.5432,
+      "step": 1001
+    },
+    {
+      "epoch": 0.09751824817518248,
+      "grad_norm": 1.2323791206307224,
+      "learning_rate": 9.881167117906276e-06,
+      "loss": 0.3361,
+      "step": 1002
+    },
+    {
+      "epoch": 0.09761557177615572,
+      "grad_norm": 1.6720448345305876,
+      "learning_rate": 9.880825288316367e-06,
+      "loss": 0.3583,
+      "step": 1003
+    },
+    {
+      "epoch": 0.09771289537712895,
+      "grad_norm": 1.408364549926656,
+      "learning_rate": 9.880482973715846e-06,
+      "loss": 0.3847,
+      "step": 1004
+    },
+    {
+      "epoch": 0.09781021897810219,
+      "grad_norm": 1.493256031544701,
+      "learning_rate": 9.880140174138735e-06,
+      "loss": 0.3611,
+      "step": 1005
+    },
+    {
+      "epoch": 0.09790754257907543,
+      "grad_norm": 1.3658283125944337,
+      "learning_rate": 9.879796889619093e-06,
+      "loss": 0.3555,
+      "step": 1006
+    },
+    {
+      "epoch": 0.09800486618004867,
+      "grad_norm": 1.7346143127846696,
+      "learning_rate": 9.879453120191037e-06,
+      "loss": 0.5028,
+      "step": 1007
+    },
+    {
+      "epoch": 0.09810218978102189,
+      "grad_norm": 1.9094090784905724,
+      "learning_rate": 9.879108865888724e-06,
+      "loss": 0.4799,
+      "step": 1008
+    },
+    {
+      "epoch": 0.09819951338199513,
+      "grad_norm": 1.1235415223499565,
+      "learning_rate": 9.878764126746364e-06,
+      "loss": 0.2181,
+      "step": 1009
+    },
+    {
+      "epoch": 0.09829683698296837,
+      "grad_norm": 1.494557121918356,
+      "learning_rate": 9.878418902798215e-06,
+      "loss": 0.4548,
+      "step": 1010
+    },
+    {
+      "epoch": 0.09839416058394161,
+      "grad_norm": 1.5340021274706077,
+      "learning_rate": 9.87807319407858e-06,
+      "loss": 0.4952,
+      "step": 1011
+    },
+    {
+      "epoch": 0.09849148418491484,
+      "grad_norm": 1.2523545024978981,
+      "learning_rate": 9.877727000621815e-06,
+      "loss": 0.2887,
+      "step": 1012
+    },
+    {
+      "epoch": 0.09858880778588808,
+      "grad_norm": 1.424446798325285,
+      "learning_rate": 9.877380322462317e-06,
+      "loss": 0.3628,
+      "step": 1013
+    },
+    {
+      "epoch": 0.09868613138686132,
+      "grad_norm": 1.6382574528105933,
+      "learning_rate": 9.877033159634542e-06,
+      "loss": 0.5396,
+      "step": 1014
+    },
+    {
+      "epoch": 0.09878345498783454,
+      "grad_norm": 1.544256440771578,
+      "learning_rate": 9.876685512172982e-06,
+      "loss": 0.4031,
+      "step": 1015
+    },
+    {
+      "epoch": 0.09888077858880778,
+      "grad_norm": 1.620162733287423,
+      "learning_rate": 9.876337380112185e-06,
+      "loss": 0.4925,
+      "step": 1016
+    },
+    {
+      "epoch": 0.09897810218978102,
+      "grad_norm": 1.6140460771461889,
+      "learning_rate": 9.875988763486746e-06,
+      "loss": 0.5549,
+      "step": 1017
+    },
+    {
+      "epoch": 0.09907542579075426,
+      "grad_norm": 1.6187864498320685,
+      "learning_rate": 9.875639662331307e-06,
+      "loss": 0.5034,
+      "step": 1018
+    },
+    {
+      "epoch": 0.09917274939172749,
+      "grad_norm": 1.249422512171971,
+      "learning_rate": 9.875290076680557e-06,
+      "loss": 0.236,
+      "step": 1019
+    },
+    {
+      "epoch": 0.09927007299270073,
+      "grad_norm": 1.5835572971087337,
+      "learning_rate": 9.874940006569236e-06,
+      "loss": 0.5309,
+      "step": 1020
+    },
+    {
+      "epoch": 0.09936739659367397,
+      "grad_norm": 0.8658795502351594,
+      "learning_rate": 9.874589452032131e-06,
+      "loss": 0.1911,
+      "step": 1021
+    },
+    {
+      "epoch": 0.09946472019464721,
+      "grad_norm": 1.3171385587421753,
+      "learning_rate": 9.874238413104076e-06,
+      "loss": 0.3486,
+      "step": 1022
+    },
+    {
+      "epoch": 0.09956204379562043,
+      "grad_norm": 1.4498439375980756,
+      "learning_rate": 9.873886889819953e-06,
+      "loss": 0.1986,
+      "step": 1023
+    },
+    {
+      "epoch": 0.09965936739659367,
+      "grad_norm": 1.5991307847988792,
+      "learning_rate": 9.873534882214692e-06,
+      "loss": 0.6397,
+      "step": 1024
+    },
+    {
+      "epoch": 0.09975669099756691,
+      "grad_norm": 1.6135151765084201,
+      "learning_rate": 9.873182390323277e-06,
+      "loss": 0.4338,
+      "step": 1025
+    },
+    {
+      "epoch": 0.09985401459854015,
+      "grad_norm": 1.465261170994732,
+      "learning_rate": 9.872829414180733e-06,
+      "loss": 0.4692,
+      "step": 1026
+    },
+    {
+      "epoch": 0.09995133819951338,
+      "grad_norm": 1.6964068418559575,
+      "learning_rate": 9.872475953822134e-06,
+      "loss": 0.4763,
+      "step": 1027
+    },
+    {
+      "epoch": 0.10004866180048662,
+      "grad_norm": 1.5209137969308788,
+      "learning_rate": 9.872122009282604e-06,
+      "loss": 0.4266,
+      "step": 1028
+    },
+    {
+      "epoch": 0.10014598540145986,
+      "grad_norm": 1.4495568716439686,
+      "learning_rate": 9.871767580597316e-06,
+      "loss": 0.4087,
+      "step": 1029
+    },
+    {
+      "epoch": 0.10024330900243308,
+      "grad_norm": 1.344434785457905,
+      "learning_rate": 9.871412667801488e-06,
+      "loss": 0.3797,
+      "step": 1030
+    },
+    {
+      "epoch": 0.10034063260340632,
+      "grad_norm": 1.5794908259633444,
+      "learning_rate": 9.871057270930392e-06,
+      "loss": 0.3939,
+      "step": 1031
+    },
+    {
+      "epoch": 0.10043795620437956,
+      "grad_norm": 1.5876979734473795,
+      "learning_rate": 9.870701390019337e-06,
+      "loss": 0.484,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1005352798053528,
+      "grad_norm": 1.8773231101994967,
+      "learning_rate": 9.870345025103694e-06,
+      "loss": 0.5893,
+      "step": 1033
+    },
+    {
+      "epoch": 0.10063260340632603,
+      "grad_norm": 1.4927383125242464,
+      "learning_rate": 9.869988176218871e-06,
+      "loss": 0.4138,
+      "step": 1034
+    },
+    {
+      "epoch": 0.10072992700729927,
+      "grad_norm": 1.4766306382054422,
+      "learning_rate": 9.869630843400331e-06,
+      "loss": 0.4125,
+      "step": 1035
+    },
+    {
+      "epoch": 0.10082725060827251,
+      "grad_norm": 2.1872385141217388,
+      "learning_rate": 9.86927302668358e-06,
+      "loss": 0.4581,
+      "step": 1036
+    },
+    {
+      "epoch": 0.10092457420924575,
+      "grad_norm": 1.4275090865666056,
+      "learning_rate": 9.868914726104174e-06,
+      "loss": 0.2393,
+      "step": 1037
+    },
+    {
+      "epoch": 0.10102189781021897,
+      "grad_norm": 1.6989614006808447,
+      "learning_rate": 9.868555941697721e-06,
+      "loss": 0.4941,
+      "step": 1038
+    },
+    {
+      "epoch": 0.10111922141119221,
+      "grad_norm": 1.4357333730365565,
+      "learning_rate": 9.86819667349987e-06,
+      "loss": 0.4907,
+      "step": 1039
+    },
+    {
+      "epoch": 0.10121654501216545,
+      "grad_norm": 2.0026376735495055,
+      "learning_rate": 9.867836921546326e-06,
+      "loss": 0.8695,
+      "step": 1040
+    },
+    {
+      "epoch": 0.10131386861313868,
+      "grad_norm": 1.6951372609783342,
+      "learning_rate": 9.867476685872833e-06,
+      "loss": 0.6236,
+      "step": 1041
+    },
+    {
+      "epoch": 0.10141119221411192,
+      "grad_norm": 1.6963236381946833,
+      "learning_rate": 9.86711596651519e-06,
+      "loss": 0.6358,
+      "step": 1042
+    },
+    {
+      "epoch": 0.10150851581508516,
+      "grad_norm": 1.5189733584329748,
+      "learning_rate": 9.866754763509242e-06,
+      "loss": 0.4374,
+      "step": 1043
+    },
+    {
+      "epoch": 0.1016058394160584,
+      "grad_norm": 1.2748045341406278,
+      "learning_rate": 9.866393076890881e-06,
+      "loss": 0.4213,
+      "step": 1044
+    },
+    {
+      "epoch": 0.10170316301703163,
+      "grad_norm": 1.7405552081322075,
+      "learning_rate": 9.866030906696051e-06,
+      "loss": 0.6708,
+      "step": 1045
+    },
+    {
+      "epoch": 0.10180048661800487,
+      "grad_norm": 1.3495682131815454,
+      "learning_rate": 9.865668252960737e-06,
+      "loss": 0.3531,
+      "step": 1046
+    },
+    {
+      "epoch": 0.1018978102189781,
+      "grad_norm": 1.5653185028552046,
+      "learning_rate": 9.86530511572098e-06,
+      "loss": 0.4331,
+      "step": 1047
+    },
+    {
+      "epoch": 0.10199513381995134,
+      "grad_norm": 1.3992858529840162,
+      "learning_rate": 9.864941495012861e-06,
+      "loss": 0.3388,
+      "step": 1048
+    },
+    {
+      "epoch": 0.10209245742092457,
+      "grad_norm": 1.6270586325333123,
+      "learning_rate": 9.864577390872516e-06,
+      "loss": 0.4234,
+      "step": 1049
+    },
+    {
+      "epoch": 0.10218978102189781,
+      "grad_norm": 1.8656971621974168,
+      "learning_rate": 9.864212803336126e-06,
+      "loss": 0.718,
+      "step": 1050
+    },
+    {
+      "epoch": 0.10228710462287105,
+      "grad_norm": 1.4029758909387644,
+      "learning_rate": 9.86384773243992e-06,
+      "loss": 0.3892,
+      "step": 1051
+    },
+    {
+      "epoch": 0.10238442822384428,
+      "grad_norm": 1.1023559958942302,
+      "learning_rate": 9.863482178220176e-06,
+      "loss": 0.2453,
+      "step": 1052
+    },
+    {
+      "epoch": 0.10248175182481752,
+      "grad_norm": 1.5775869982106272,
+      "learning_rate": 9.863116140713219e-06,
+      "loss": 0.5324,
+      "step": 1053
+    },
+    {
+      "epoch": 0.10257907542579076,
+      "grad_norm": 1.603675899324949,
+      "learning_rate": 9.86274961995542e-06,
+      "loss": 0.4521,
+      "step": 1054
+    },
+    {
+      "epoch": 0.102676399026764,
+      "grad_norm": 1.6020699046167006,
+      "learning_rate": 9.862382615983203e-06,
+      "loss": 0.4545,
+      "step": 1055
+    },
+    {
+      "epoch": 0.10277372262773722,
+      "grad_norm": 1.474718021659803,
+      "learning_rate": 9.862015128833036e-06,
+      "loss": 0.4822,
+      "step": 1056
+    },
+    {
+      "epoch": 0.10287104622871046,
+      "grad_norm": 1.6033514684549,
+      "learning_rate": 9.861647158541438e-06,
+      "loss": 0.5069,
+      "step": 1057
+    },
+    {
+      "epoch": 0.1029683698296837,
+      "grad_norm": 1.4841655382640788,
+      "learning_rate": 9.861278705144974e-06,
+      "loss": 0.3865,
+      "step": 1058
+    },
+    {
+      "epoch": 0.10306569343065694,
+      "grad_norm": 1.1425556408878823,
+      "learning_rate": 9.860909768680259e-06,
+      "loss": 0.2443,
+      "step": 1059
+    },
+    {
+      "epoch": 0.10316301703163017,
+      "grad_norm": 1.5288676978753954,
+      "learning_rate": 9.86054034918395e-06,
+      "loss": 0.3652,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1032603406326034,
+      "grad_norm": 1.5264484093473076,
+      "learning_rate": 9.860170446692758e-06,
+      "loss": 0.3318,
+      "step": 1061
+    },
+    {
+      "epoch": 0.10335766423357665,
+      "grad_norm": 1.4476258605632986,
+      "learning_rate": 9.859800061243443e-06,
+      "loss": 0.4518,
+      "step": 1062
+    },
+    {
+      "epoch": 0.10345498783454987,
+      "grad_norm": 1.336933590040686,
+      "learning_rate": 9.859429192872809e-06,
+      "loss": 0.2652,
+      "step": 1063
+    },
+    {
+      "epoch": 0.10355231143552311,
+      "grad_norm": 1.6050187197155075,
+      "learning_rate": 9.859057841617709e-06,
+      "loss": 0.5383,
+      "step": 1064
+    },
+    {
+      "epoch": 0.10364963503649635,
+      "grad_norm": 1.3472405276196469,
+      "learning_rate": 9.858686007515045e-06,
+      "loss": 0.4483,
+      "step": 1065
+    },
+    {
+      "epoch": 0.10374695863746959,
+      "grad_norm": 1.4838970820374793,
+      "learning_rate": 9.858313690601767e-06,
+      "loss": 0.3506,
+      "step": 1066
+    },
+    {
+      "epoch": 0.10384428223844282,
+      "grad_norm": 1.5911831099601979,
+      "learning_rate": 9.857940890914868e-06,
+      "loss": 0.3995,
+      "step": 1067
+    },
+    {
+      "epoch": 0.10394160583941606,
+      "grad_norm": 1.415577451063168,
+      "learning_rate": 9.8575676084914e-06,
+      "loss": 0.4773,
+      "step": 1068
+    },
+    {
+      "epoch": 0.1040389294403893,
+      "grad_norm": 1.7250253730787564,
+      "learning_rate": 9.857193843368451e-06,
+      "loss": 0.4456,
+      "step": 1069
+    },
+    {
+      "epoch": 0.10413625304136254,
+      "grad_norm": 1.5066269873708278,
+      "learning_rate": 9.856819595583166e-06,
+      "loss": 0.5481,
+      "step": 1070
+    },
+    {
+      "epoch": 0.10423357664233576,
+      "grad_norm": 1.5626665408071483,
+      "learning_rate": 9.856444865172732e-06,
+      "loss": 0.5382,
+      "step": 1071
+    },
+    {
+      "epoch": 0.104330900243309,
+      "grad_norm": 1.9089561390061884,
+      "learning_rate": 9.856069652174385e-06,
+      "loss": 0.5533,
+      "step": 1072
+    },
+    {
+      "epoch": 0.10442822384428224,
+      "grad_norm": 1.2757688373398666,
+      "learning_rate": 9.855693956625414e-06,
+      "loss": 0.3065,
+      "step": 1073
+    },
+    {
+      "epoch": 0.10452554744525547,
+      "grad_norm": 1.7230598513214688,
+      "learning_rate": 9.85531777856315e-06,
+      "loss": 0.5367,
+      "step": 1074
+    },
+    {
+      "epoch": 0.10462287104622871,
+      "grad_norm": 1.8368494244508635,
+      "learning_rate": 9.854941118024973e-06,
+      "loss": 0.4587,
+      "step": 1075
+    },
+    {
+      "epoch": 0.10472019464720195,
+      "grad_norm": 1.418583003899538,
+      "learning_rate": 9.854563975048314e-06,
+      "loss": 0.405,
+      "step": 1076
+    },
+    {
+      "epoch": 0.10481751824817519,
+      "grad_norm": 1.555078045275604,
+      "learning_rate": 9.854186349670648e-06,
+      "loss": 0.5572,
+      "step": 1077
+    },
+    {
+      "epoch": 0.10491484184914841,
+      "grad_norm": 1.5414220083120458,
+      "learning_rate": 9.853808241929502e-06,
+      "loss": 0.3382,
+      "step": 1078
+    },
+    {
+      "epoch": 0.10501216545012165,
+      "grad_norm": 1.2895897451723073,
+      "learning_rate": 9.853429651862445e-06,
+      "loss": 0.4342,
+      "step": 1079
+    },
+    {
+      "epoch": 0.10510948905109489,
+      "grad_norm": 1.3117010773132232,
+      "learning_rate": 9.853050579507104e-06,
+      "loss": 0.3751,
+      "step": 1080
+    },
+    {
+      "epoch": 0.10520681265206813,
+      "grad_norm": 1.5440994948167002,
+      "learning_rate": 9.852671024901141e-06,
+      "loss": 0.4971,
+      "step": 1081
+    },
+    {
+      "epoch": 0.10530413625304136,
+      "grad_norm": 1.2028388141262132,
+      "learning_rate": 9.852290988082278e-06,
+      "loss": 0.3933,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1054014598540146,
+      "grad_norm": 1.6199890049219825,
+      "learning_rate": 9.851910469088275e-06,
+      "loss": 0.5394,
+      "step": 1083
+    },
+    {
+      "epoch": 0.10549878345498784,
+      "grad_norm": 1.4805170620003079,
+      "learning_rate": 9.851529467956946e-06,
+      "loss": 0.2421,
+      "step": 1084
+    },
+    {
+      "epoch": 0.10559610705596106,
+      "grad_norm": 1.432802486072686,
+      "learning_rate": 9.851147984726154e-06,
+      "loss": 0.479,
+      "step": 1085
+    },
+    {
+      "epoch": 0.1056934306569343,
+      "grad_norm": 1.7662999036343905,
+      "learning_rate": 9.850766019433803e-06,
+      "loss": 0.706,
+      "step": 1086
+    },
+    {
+      "epoch": 0.10579075425790754,
+      "grad_norm": 1.9136497208168854,
+      "learning_rate": 9.850383572117853e-06,
+      "loss": 0.7672,
+      "step": 1087
+    },
+    {
+      "epoch": 0.10588807785888078,
+      "grad_norm": 1.1667281997438979,
+      "learning_rate": 9.850000642816306e-06,
+      "loss": 0.2263,
+      "step": 1088
+    },
+    {
+      "epoch": 0.10598540145985401,
+      "grad_norm": 1.3133144576431575,
+      "learning_rate": 9.849617231567213e-06,
+      "loss": 0.2211,
+      "step": 1089
+    },
+    {
+      "epoch": 0.10608272506082725,
+      "grad_norm": 1.411642205718121,
+      "learning_rate": 9.849233338408674e-06,
+      "loss": 0.4379,
+      "step": 1090
+    },
+    {
+      "epoch": 0.10618004866180049,
+      "grad_norm": 1.7114110143353651,
+      "learning_rate": 9.84884896337884e-06,
+      "loss": 0.462,
+      "step": 1091
+    },
+    {
+      "epoch": 0.10627737226277373,
+      "grad_norm": 1.4035875335457177,
+      "learning_rate": 9.848464106515903e-06,
+      "loss": 0.317,
+      "step": 1092
+    },
+    {
+      "epoch": 0.10637469586374695,
+      "grad_norm": 1.5988244446936477,
+      "learning_rate": 9.848078767858107e-06,
+      "loss": 0.5254,
+      "step": 1093
+    },
+    {
+      "epoch": 0.1064720194647202,
+      "grad_norm": 1.6336010940510732,
+      "learning_rate": 9.847692947443745e-06,
+      "loss": 0.4979,
+      "step": 1094
+    },
+    {
+      "epoch": 0.10656934306569343,
+      "grad_norm": 1.68747146017171,
+      "learning_rate": 9.847306645311154e-06,
+      "loss": 0.5515,
+      "step": 1095
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 1.497709273552353,
+      "learning_rate": 9.846919861498724e-06,
+      "loss": 0.4221,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1067639902676399,
+      "grad_norm": 1.4761873606313476,
+      "learning_rate": 9.846532596044887e-06,
+      "loss": 0.4296,
+      "step": 1097
+    },
+    {
+      "epoch": 0.10686131386861314,
+      "grad_norm": 1.1441862868877024,
+      "learning_rate": 9.846144848988127e-06,
+      "loss": 0.2816,
+      "step": 1098
+    },
+    {
+      "epoch": 0.10695863746958638,
+      "grad_norm": 1.7272604657837642,
+      "learning_rate": 9.845756620366976e-06,
+      "loss": 0.5916,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1070559610705596,
+      "grad_norm": 1.3799505412872324,
+      "learning_rate": 9.84536791022001e-06,
+      "loss": 0.3947,
+      "step": 1100
+    },
+    {
+      "epoch": 0.10715328467153284,
+      "grad_norm": 1.6943818099878132,
+      "learning_rate": 9.844978718585855e-06,
+      "loss": 0.4737,
+      "step": 1101
+    },
+    {
+      "epoch": 0.10725060827250608,
+      "grad_norm": 1.5405614688920448,
+      "learning_rate": 9.84458904550319e-06,
+      "loss": 0.4152,
+      "step": 1102
+    },
+    {
+      "epoch": 0.10734793187347932,
+      "grad_norm": 1.6335292867295117,
+      "learning_rate": 9.844198891010733e-06,
+      "loss": 0.5677,
+      "step": 1103
+    },
+    {
+      "epoch": 0.10744525547445255,
+      "grad_norm": 1.302603147379972,
+      "learning_rate": 9.843808255147253e-06,
+      "loss": 0.4283,
+      "step": 1104
+    },
+    {
+      "epoch": 0.10754257907542579,
+      "grad_norm": 1.7967506033919078,
+      "learning_rate": 9.84341713795157e-06,
+      "loss": 0.6995,
+      "step": 1105
+    },
+    {
+      "epoch": 0.10763990267639903,
+      "grad_norm": 1.7320527346822367,
+      "learning_rate": 9.84302553946255e-06,
+      "loss": 0.5369,
+      "step": 1106
+    },
+    {
+      "epoch": 0.10773722627737227,
+      "grad_norm": 1.2124746103676287,
+      "learning_rate": 9.842633459719104e-06,
+      "loss": 0.296,
+      "step": 1107
+    },
+    {
+      "epoch": 0.1078345498783455,
+      "grad_norm": 1.6638227119864655,
+      "learning_rate": 9.842240898760195e-06,
+      "loss": 0.5632,
+      "step": 1108
+    },
+    {
+      "epoch": 0.10793187347931874,
+      "grad_norm": 1.5728826792836543,
+      "learning_rate": 9.841847856624833e-06,
+      "loss": 0.3407,
+      "step": 1109
+    },
+    {
+      "epoch": 0.10802919708029197,
+      "grad_norm": 1.4855225795030034,
+      "learning_rate": 9.841454333352073e-06,
+      "loss": 0.534,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1081265206812652,
+      "grad_norm": 1.741747608159628,
+      "learning_rate": 9.841060328981019e-06,
+      "loss": 0.5739,
+      "step": 1111
+    },
+    {
+      "epoch": 0.10822384428223844,
+      "grad_norm": 1.2765533148109443,
+      "learning_rate": 9.840665843550825e-06,
+      "loss": 0.335,
+      "step": 1112
+    },
+    {
+      "epoch": 0.10832116788321168,
+      "grad_norm": 1.9391527817309226,
+      "learning_rate": 9.840270877100692e-06,
+      "loss": 0.5604,
+      "step": 1113
+    },
+    {
+      "epoch": 0.10841849148418492,
+      "grad_norm": 1.2570937099076989,
+      "learning_rate": 9.839875429669865e-06,
+      "loss": 0.3098,
+      "step": 1114
+    },
+    {
+      "epoch": 0.10851581508515815,
+      "grad_norm": 1.6345857910998665,
+      "learning_rate": 9.839479501297643e-06,
+      "loss": 0.4665,
+      "step": 1115
+    },
+    {
+      "epoch": 0.10861313868613139,
+      "grad_norm": 2.1039943309751075,
+      "learning_rate": 9.839083092023368e-06,
+      "loss": 0.8597,
+      "step": 1116
+    },
+    {
+      "epoch": 0.10871046228710463,
+      "grad_norm": 1.634678554608885,
+      "learning_rate": 9.838686201886432e-06,
+      "loss": 0.4907,
+      "step": 1117
+    },
+    {
+      "epoch": 0.10880778588807787,
+      "grad_norm": 1.328229383966676,
+      "learning_rate": 9.838288830926274e-06,
+      "loss": 0.3255,
+      "step": 1118
+    },
+    {
+      "epoch": 0.10890510948905109,
+      "grad_norm": 1.3587359099021656,
+      "learning_rate": 9.837890979182381e-06,
+      "loss": 0.4224,
+      "step": 1119
+    },
+    {
+      "epoch": 0.10900243309002433,
+      "grad_norm": 1.6242900911620413,
+      "learning_rate": 9.837492646694287e-06,
+      "loss": 0.4338,
+      "step": 1120
+    },
+    {
+      "epoch": 0.10909975669099757,
+      "grad_norm": 1.5901048900387273,
+      "learning_rate": 9.837093833501576e-06,
+      "loss": 0.5168,
+      "step": 1121
+    },
+    {
+      "epoch": 0.1091970802919708,
+      "grad_norm": 1.34172908606168,
+      "learning_rate": 9.836694539643878e-06,
+      "loss": 0.3233,
+      "step": 1122
+    },
+    {
+      "epoch": 0.10929440389294404,
+      "grad_norm": 1.4724714330159256,
+      "learning_rate": 9.83629476516087e-06,
+      "loss": 0.3652,
+      "step": 1123
+    },
+    {
+      "epoch": 0.10939172749391728,
+      "grad_norm": 1.4884050773310515,
+      "learning_rate": 9.835894510092279e-06,
+      "loss": 0.4622,
+      "step": 1124
+    },
+    {
+      "epoch": 0.10948905109489052,
+      "grad_norm": 1.3181328020728609,
+      "learning_rate": 9.835493774477877e-06,
+      "loss": 0.4531,
+      "step": 1125
+    },
+    {
+      "epoch": 0.10958637469586374,
+      "grad_norm": 1.5414298966880746,
+      "learning_rate": 9.835092558357488e-06,
+      "loss": 0.3659,
+      "step": 1126
+    },
+    {
+      "epoch": 0.10968369829683698,
+      "grad_norm": 1.3248299507567909,
+      "learning_rate": 9.834690861770979e-06,
+      "loss": 0.3207,
+      "step": 1127
+    },
+    {
+      "epoch": 0.10978102189781022,
+      "grad_norm": 1.5527535683267375,
+      "learning_rate": 9.834288684758269e-06,
+      "loss": 0.4938,
+      "step": 1128
+    },
+    {
+      "epoch": 0.10987834549878346,
+      "grad_norm": 1.3342131255187983,
+      "learning_rate": 9.83388602735932e-06,
+      "loss": 0.4451,
+      "step": 1129
+    },
+    {
+      "epoch": 0.10997566909975669,
+      "grad_norm": 1.0500905202266426,
+      "learning_rate": 9.833482889614143e-06,
+      "loss": 0.2408,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11007299270072993,
+      "grad_norm": 1.377353907486564,
+      "learning_rate": 9.833079271562802e-06,
+      "loss": 0.3945,
+      "step": 1131
+    },
+    {
+      "epoch": 0.11017031630170317,
+      "grad_norm": 1.5823324787969848,
+      "learning_rate": 9.832675173245404e-06,
+      "loss": 0.6066,
+      "step": 1132
+    },
+    {
+      "epoch": 0.11026763990267639,
+      "grad_norm": 1.7266167679625446,
+      "learning_rate": 9.832270594702102e-06,
+      "loss": 0.6417,
+      "step": 1133
+    },
+    {
+      "epoch": 0.11036496350364963,
+      "grad_norm": 1.4091165783577269,
+      "learning_rate": 9.831865535973103e-06,
+      "loss": 0.2661,
+      "step": 1134
+    },
+    {
+      "epoch": 0.11046228710462287,
+      "grad_norm": 0.9959339686876645,
+      "learning_rate": 9.831459997098654e-06,
+      "loss": 0.1744,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11055961070559611,
+      "grad_norm": 1.4748243970921762,
+      "learning_rate": 9.831053978119056e-06,
+      "loss": 0.4011,
+      "step": 1136
+    },
+    {
+      "epoch": 0.11065693430656934,
+      "grad_norm": 1.5879686249629044,
+      "learning_rate": 9.830647479074656e-06,
+      "loss": 0.3021,
+      "step": 1137
+    },
+    {
+      "epoch": 0.11075425790754258,
+      "grad_norm": 1.5057704716227702,
+      "learning_rate": 9.830240500005845e-06,
+      "loss": 0.2962,
+      "step": 1138
+    },
+    {
+      "epoch": 0.11085158150851582,
+      "grad_norm": 1.7497051535586357,
+      "learning_rate": 9.829833040953068e-06,
+      "loss": 0.4717,
+      "step": 1139
+    },
+    {
+      "epoch": 0.11094890510948906,
+      "grad_norm": 1.7819946472609902,
+      "learning_rate": 9.829425101956812e-06,
+      "loss": 0.6113,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11104622871046228,
+      "grad_norm": 1.7680522472506797,
+      "learning_rate": 9.829016683057615e-06,
+      "loss": 0.4672,
+      "step": 1141
+    },
+    {
+      "epoch": 0.11114355231143552,
+      "grad_norm": 1.8291787265156998,
+      "learning_rate": 9.828607784296063e-06,
+      "loss": 0.5148,
+      "step": 1142
+    },
+    {
+      "epoch": 0.11124087591240876,
+      "grad_norm": 1.4119536127948566,
+      "learning_rate": 9.828198405712788e-06,
+      "loss": 0.2698,
+      "step": 1143
+    },
+    {
+      "epoch": 0.11133819951338199,
+      "grad_norm": 1.67600232780131,
+      "learning_rate": 9.827788547348469e-06,
+      "loss": 0.4912,
+      "step": 1144
+    },
+    {
+      "epoch": 0.11143552311435523,
+      "grad_norm": 1.9367616538665617,
+      "learning_rate": 9.827378209243835e-06,
+      "loss": 0.3781,
+      "step": 1145
+    },
+    {
+      "epoch": 0.11153284671532847,
+      "grad_norm": 1.7032208896905794,
+      "learning_rate": 9.826967391439662e-06,
+      "loss": 0.5816,
+      "step": 1146
+    },
+    {
+      "epoch": 0.11163017031630171,
+      "grad_norm": 1.60872896431165,
+      "learning_rate": 9.826556093976769e-06,
+      "loss": 0.4654,
+      "step": 1147
+    },
+    {
+      "epoch": 0.11172749391727493,
+      "grad_norm": 1.5752275514466696,
+      "learning_rate": 9.826144316896033e-06,
+      "loss": 0.3177,
+      "step": 1148
+    },
+    {
+      "epoch": 0.11182481751824817,
+      "grad_norm": 1.8207599924827627,
+      "learning_rate": 9.82573206023837e-06,
+      "loss": 0.5701,
+      "step": 1149
+    },
+    {
+      "epoch": 0.11192214111922141,
+      "grad_norm": 1.5850279506541385,
+      "learning_rate": 9.825319324044745e-06,
+      "loss": 0.5616,
+      "step": 1150
+    },
+    {
+      "epoch": 0.11201946472019465,
+      "grad_norm": 1.360496233978723,
+      "learning_rate": 9.824906108356174e-06,
+      "loss": 0.3407,
+      "step": 1151
+    },
+    {
+      "epoch": 0.11211678832116788,
+      "grad_norm": 1.6595565610362801,
+      "learning_rate": 9.824492413213717e-06,
+      "loss": 0.6641,
+      "step": 1152
+    },
+    {
+      "epoch": 0.11221411192214112,
+      "grad_norm": 1.6031792644515102,
+      "learning_rate": 9.824078238658483e-06,
+      "loss": 0.4779,
+      "step": 1153
+    },
+    {
+      "epoch": 0.11231143552311436,
+      "grad_norm": 1.0762751645680708,
+      "learning_rate": 9.82366358473163e-06,
+      "loss": 0.2739,
+      "step": 1154
+    },
+    {
+      "epoch": 0.11240875912408758,
+      "grad_norm": 1.3660129842713564,
+      "learning_rate": 9.82324845147436e-06,
+      "loss": 0.5043,
+      "step": 1155
+    },
+    {
+      "epoch": 0.11250608272506082,
+      "grad_norm": 1.6273408315616833,
+      "learning_rate": 9.822832838927929e-06,
+      "loss": 0.6159,
+      "step": 1156
+    },
+    {
+      "epoch": 0.11260340632603406,
+      "grad_norm": 1.4216921342906768,
+      "learning_rate": 9.822416747133634e-06,
+      "loss": 0.4093,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1127007299270073,
+      "grad_norm": 1.8899721642114575,
+      "learning_rate": 9.822000176132822e-06,
+      "loss": 0.5586,
+      "step": 1158
+    },
+    {
+      "epoch": 0.11279805352798053,
+      "grad_norm": 1.5144459966059345,
+      "learning_rate": 9.821583125966889e-06,
+      "loss": 0.3806,
+      "step": 1159
+    },
+    {
+      "epoch": 0.11289537712895377,
+      "grad_norm": 1.61041803725934,
+      "learning_rate": 9.821165596677278e-06,
+      "loss": 0.4064,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11299270072992701,
+      "grad_norm": 1.5410637406837986,
+      "learning_rate": 9.820747588305477e-06,
+      "loss": 0.3526,
+      "step": 1161
+    },
+    {
+      "epoch": 0.11309002433090025,
+      "grad_norm": 1.5545393523360629,
+      "learning_rate": 9.820329100893026e-06,
+      "loss": 0.3834,
+      "step": 1162
+    },
+    {
+      "epoch": 0.11318734793187347,
+      "grad_norm": 1.6391567381322345,
+      "learning_rate": 9.819910134481508e-06,
+      "loss": 0.3849,
+      "step": 1163
+    },
+    {
+      "epoch": 0.11328467153284671,
+      "grad_norm": 1.5204183543600032,
+      "learning_rate": 9.819490689112559e-06,
+      "loss": 0.4712,
+      "step": 1164
+    },
+    {
+      "epoch": 0.11338199513381995,
+      "grad_norm": 1.5168954302933022,
+      "learning_rate": 9.819070764827858e-06,
+      "loss": 0.4662,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1134793187347932,
+      "grad_norm": 1.4412304117107342,
+      "learning_rate": 9.818650361669133e-06,
+      "loss": 0.3515,
+      "step": 1166
+    },
+    {
+      "epoch": 0.11357664233576642,
+      "grad_norm": 1.5419710047923603,
+      "learning_rate": 9.81822947967816e-06,
+      "loss": 0.383,
+      "step": 1167
+    },
+    {
+      "epoch": 0.11367396593673966,
+      "grad_norm": 1.59211707141906,
+      "learning_rate": 9.817808118896759e-06,
+      "loss": 0.5101,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1137712895377129,
+      "grad_norm": 1.9315831066859817,
+      "learning_rate": 9.817386279366808e-06,
+      "loss": 0.6179,
+      "step": 1169
+    },
+    {
+      "epoch": 0.11386861313868613,
+      "grad_norm": 1.3153157684002792,
+      "learning_rate": 9.816963961130218e-06,
+      "loss": 0.2382,
+      "step": 1170
+    },
+    {
+      "epoch": 0.11396593673965937,
+      "grad_norm": 1.3579619945410324,
+      "learning_rate": 9.81654116422896e-06,
+      "loss": 0.4424,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1140632603406326,
+      "grad_norm": 1.479330223962703,
+      "learning_rate": 9.816117888705046e-06,
+      "loss": 0.3647,
+      "step": 1172
+    },
+    {
+      "epoch": 0.11416058394160584,
+      "grad_norm": 1.5031676224913018,
+      "learning_rate": 9.815694134600537e-06,
+      "loss": 0.3686,
+      "step": 1173
+    },
+    {
+      "epoch": 0.11425790754257907,
+      "grad_norm": 1.6106095254885215,
+      "learning_rate": 9.815269901957543e-06,
+      "loss": 0.5309,
+      "step": 1174
+    },
+    {
+      "epoch": 0.11435523114355231,
+      "grad_norm": 1.4367590943688036,
+      "learning_rate": 9.814845190818218e-06,
+      "loss": 0.3786,
+      "step": 1175
+    },
+    {
+      "epoch": 0.11445255474452555,
+      "grad_norm": 2.0513510648109636,
+      "learning_rate": 9.814420001224767e-06,
+      "loss": 0.8885,
+      "step": 1176
+    },
+    {
+      "epoch": 0.11454987834549879,
+      "grad_norm": 1.3799990465326748,
+      "learning_rate": 9.813994333219443e-06,
+      "loss": 0.3511,
+      "step": 1177
+    },
+    {
+      "epoch": 0.11464720194647202,
+      "grad_norm": 1.2354207015762353,
+      "learning_rate": 9.813568186844541e-06,
+      "loss": 0.3571,
+      "step": 1178
+    },
+    {
+      "epoch": 0.11474452554744526,
+      "grad_norm": 2.0501383618438678,
+      "learning_rate": 9.813141562142409e-06,
+      "loss": 0.4485,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1148418491484185,
+      "grad_norm": 1.351584991091541,
+      "learning_rate": 9.812714459155444e-06,
+      "loss": 0.2894,
+      "step": 1180
+    },
+    {
+      "epoch": 0.11493917274939172,
+      "grad_norm": 1.3568994189032655,
+      "learning_rate": 9.812286877926085e-06,
+      "loss": 0.4016,
+      "step": 1181
+    },
+    {
+      "epoch": 0.11503649635036496,
+      "grad_norm": 1.4949546840268106,
+      "learning_rate": 9.81185881849682e-06,
+      "loss": 0.527,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1151338199513382,
+      "grad_norm": 1.5053242129518953,
+      "learning_rate": 9.811430280910186e-06,
+      "loss": 0.4324,
+      "step": 1183
+    },
+    {
+      "epoch": 0.11523114355231144,
+      "grad_norm": 1.2995408017430223,
+      "learning_rate": 9.811001265208768e-06,
+      "loss": 0.4592,
+      "step": 1184
+    },
+    {
+      "epoch": 0.11532846715328467,
+      "grad_norm": 1.4103061247668216,
+      "learning_rate": 9.810571771435197e-06,
+      "loss": 0.4615,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1154257907542579,
+      "grad_norm": 1.3694132099540144,
+      "learning_rate": 9.810141799632153e-06,
+      "loss": 0.4224,
+      "step": 1186
+    },
+    {
+      "epoch": 0.11552311435523115,
+      "grad_norm": 1.4494836775882813,
+      "learning_rate": 9.809711349842363e-06,
+      "loss": 0.4189,
+      "step": 1187
+    },
+    {
+      "epoch": 0.11562043795620439,
+      "grad_norm": 1.5100099037805617,
+      "learning_rate": 9.809280422108598e-06,
+      "loss": 0.495,
+      "step": 1188
+    },
+    {
+      "epoch": 0.11571776155717761,
+      "grad_norm": 1.449093301695385,
+      "learning_rate": 9.808849016473682e-06,
+      "loss": 0.345,
+      "step": 1189
+    },
+    {
+      "epoch": 0.11581508515815085,
+      "grad_norm": 1.501093862959825,
+      "learning_rate": 9.808417132980484e-06,
+      "loss": 0.4624,
+      "step": 1190
+    },
+    {
+      "epoch": 0.11591240875912409,
+      "grad_norm": 1.4567657310588336,
+      "learning_rate": 9.807984771671919e-06,
+      "loss": 0.2836,
+      "step": 1191
+    },
+    {
+      "epoch": 0.11600973236009732,
+      "grad_norm": 1.6666134190000732,
+      "learning_rate": 9.807551932590952e-06,
+      "loss": 0.3341,
+      "step": 1192
+    },
+    {
+      "epoch": 0.11610705596107056,
+      "grad_norm": 1.7534770482902293,
+      "learning_rate": 9.807118615780595e-06,
+      "loss": 0.6021,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1162043795620438,
+      "grad_norm": 1.744738707996039,
+      "learning_rate": 9.806684821283908e-06,
+      "loss": 0.4593,
+      "step": 1194
+    },
+    {
+      "epoch": 0.11630170316301704,
+      "grad_norm": 1.7519974888996959,
+      "learning_rate": 9.806250549143994e-06,
+      "loss": 0.5433,
+      "step": 1195
+    },
+    {
+      "epoch": 0.11639902676399026,
+      "grad_norm": 1.6094009249182397,
+      "learning_rate": 9.805815799404008e-06,
+      "loss": 0.6053,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1164963503649635,
+      "grad_norm": 1.4291146386614342,
+      "learning_rate": 9.805380572107153e-06,
+      "loss": 0.4377,
+      "step": 1197
+    },
+    {
+      "epoch": 0.11659367396593674,
+      "grad_norm": 1.6092739629047335,
+      "learning_rate": 9.804944867296678e-06,
+      "loss": 0.5708,
+      "step": 1198
+    },
+    {
+      "epoch": 0.11669099756690998,
+      "grad_norm": 1.3856208861087336,
+      "learning_rate": 9.804508685015876e-06,
+      "loss": 0.3677,
+      "step": 1199
+    },
+    {
+      "epoch": 0.11678832116788321,
+      "grad_norm": 1.52110832375871,
+      "learning_rate": 9.804072025308096e-06,
+      "loss": 0.3076,
+      "step": 1200
+    },
+    {
+      "epoch": 0.11688564476885645,
+      "grad_norm": 1.3072729020716074,
+      "learning_rate": 9.803634888216724e-06,
+      "loss": 0.2673,
+      "step": 1201
+    },
+    {
+      "epoch": 0.11698296836982969,
+      "grad_norm": 1.9045471339964295,
+      "learning_rate": 9.8031972737852e-06,
+      "loss": 0.7326,
+      "step": 1202
+    },
+    {
+      "epoch": 0.11708029197080291,
+      "grad_norm": 1.3351659498760804,
+      "learning_rate": 9.802759182057013e-06,
+      "loss": 0.4193,
+      "step": 1203
+    },
+    {
+      "epoch": 0.11717761557177615,
+      "grad_norm": 1.4664570380446003,
+      "learning_rate": 9.80232061307569e-06,
+      "loss": 0.358,
+      "step": 1204
+    },
+    {
+      "epoch": 0.11727493917274939,
+      "grad_norm": 1.1764722042212887,
+      "learning_rate": 9.80188156688482e-06,
+      "loss": 0.3093,
+      "step": 1205
+    },
+    {
+      "epoch": 0.11737226277372263,
+      "grad_norm": 1.5415184448059258,
+      "learning_rate": 9.801442043528026e-06,
+      "loss": 0.4667,
+      "step": 1206
+    },
+    {
+      "epoch": 0.11746958637469586,
+      "grad_norm": 1.4827166479100118,
+      "learning_rate": 9.801002043048984e-06,
+      "loss": 0.4876,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1175669099756691,
+      "grad_norm": 1.6786553338713377,
+      "learning_rate": 9.80056156549142e-06,
+      "loss": 0.5076,
+      "step": 1208
+    },
+    {
+      "epoch": 0.11766423357664234,
+      "grad_norm": 1.2161366736688597,
+      "learning_rate": 9.8001206108991e-06,
+      "loss": 0.2247,
+      "step": 1209
+    },
+    {
+      "epoch": 0.11776155717761558,
+      "grad_norm": 1.4015628266094937,
+      "learning_rate": 9.799679179315846e-06,
+      "loss": 0.4327,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1178588807785888,
+      "grad_norm": 1.5420255844625947,
+      "learning_rate": 9.799237270785522e-06,
+      "loss": 0.438,
+      "step": 1211
+    },
+    {
+      "epoch": 0.11795620437956204,
+      "grad_norm": 1.5978057716745744,
+      "learning_rate": 9.79879488535204e-06,
+      "loss": 0.4203,
+      "step": 1212
+    },
+    {
+      "epoch": 0.11805352798053528,
+      "grad_norm": 1.8973070083198396,
+      "learning_rate": 9.79835202305936e-06,
+      "loss": 0.7404,
+      "step": 1213
+    },
+    {
+      "epoch": 0.11815085158150851,
+      "grad_norm": 1.5331088091760856,
+      "learning_rate": 9.797908683951492e-06,
+      "loss": 0.5378,
+      "step": 1214
+    },
+    {
+      "epoch": 0.11824817518248175,
+      "grad_norm": 1.9627839775910105,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 0.6298,
+      "step": 1215
+    },
+    {
+      "epoch": 0.11834549878345499,
+      "grad_norm": 1.5059209630421948,
+      "learning_rate": 9.797020575466452e-06,
+      "loss": 0.4233,
+      "step": 1216
+    },
+    {
+      "epoch": 0.11844282238442823,
+      "grad_norm": 1.4714593450262028,
+      "learning_rate": 9.796575806177531e-06,
+      "loss": 0.4078,
+      "step": 1217
+    },
+    {
+      "epoch": 0.11854014598540145,
+      "grad_norm": 1.812199008547911,
+      "learning_rate": 9.796130560249926e-06,
+      "loss": 0.6636,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1186374695863747,
+      "grad_norm": 1.330364448549248,
+      "learning_rate": 9.795684837727878e-06,
+      "loss": 0.2597,
+      "step": 1219
+    },
+    {
+      "epoch": 0.11873479318734793,
+      "grad_norm": 1.1642089342014024,
+      "learning_rate": 9.795238638655681e-06,
+      "loss": 0.2669,
+      "step": 1220
+    },
+    {
+      "epoch": 0.11883211678832117,
+      "grad_norm": 1.0578785975666756,
+      "learning_rate": 9.794791963077672e-06,
+      "loss": 0.2138,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1189294403892944,
+      "grad_norm": 1.2810119779981208,
+      "learning_rate": 9.794344811038239e-06,
+      "loss": 0.3426,
+      "step": 1222
+    },
+    {
+      "epoch": 0.11902676399026764,
+      "grad_norm": 1.6109574325023976,
+      "learning_rate": 9.793897182581816e-06,
+      "loss": 0.4931,
+      "step": 1223
+    },
+    {
+      "epoch": 0.11912408759124088,
+      "grad_norm": 1.8314564663365431,
+      "learning_rate": 9.793449077752882e-06,
+      "loss": 0.5424,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1192214111922141,
+      "grad_norm": 1.3266514401224994,
+      "learning_rate": 9.793000496595968e-06,
+      "loss": 0.3123,
+      "step": 1225
+    },
+    {
+      "epoch": 0.11931873479318734,
+      "grad_norm": 1.624792232435884,
+      "learning_rate": 9.792551439155649e-06,
+      "loss": 0.3635,
+      "step": 1226
+    },
+    {
+      "epoch": 0.11941605839416058,
+      "grad_norm": 1.306535519875853,
+      "learning_rate": 9.792101905476547e-06,
+      "loss": 0.3252,
+      "step": 1227
+    },
+    {
+      "epoch": 0.11951338199513382,
+      "grad_norm": 1.591218471169796,
+      "learning_rate": 9.791651895603333e-06,
+      "loss": 0.5493,
+      "step": 1228
+    },
+    {
+      "epoch": 0.11961070559610705,
+      "grad_norm": 1.8218114354346657,
+      "learning_rate": 9.791201409580725e-06,
+      "loss": 0.6988,
+      "step": 1229
+    },
+    {
+      "epoch": 0.11970802919708029,
+      "grad_norm": 1.7366783724272585,
+      "learning_rate": 9.790750447453487e-06,
+      "loss": 0.4285,
+      "step": 1230
+    },
+    {
+      "epoch": 0.11980535279805353,
+      "grad_norm": 1.9439764988659998,
+      "learning_rate": 9.790299009266434e-06,
+      "loss": 0.2787,
+      "step": 1231
+    },
+    {
+      "epoch": 0.11990267639902677,
+      "grad_norm": 1.4894849660267724,
+      "learning_rate": 9.789847095064425e-06,
+      "loss": 0.2531,
+      "step": 1232
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6270936536604101,
+      "learning_rate": 9.789394704892364e-06,
+      "loss": 0.5309,
+      "step": 1233
+    },
+    {
+      "epoch": 0.12009732360097324,
+      "grad_norm": 1.4144832764840753,
+      "learning_rate": 9.788941838795209e-06,
+      "loss": 0.298,
+      "step": 1234
+    },
+    {
+      "epoch": 0.12019464720194648,
+      "grad_norm": 1.546926786538444,
+      "learning_rate": 9.788488496817958e-06,
+      "loss": 0.4751,
+      "step": 1235
+    },
+    {
+      "epoch": 0.12029197080291971,
+      "grad_norm": 1.5827216255031866,
+      "learning_rate": 9.788034679005664e-06,
+      "loss": 0.4576,
+      "step": 1236
+    },
+    {
+      "epoch": 0.12038929440389294,
+      "grad_norm": 1.6103699210596951,
+      "learning_rate": 9.78758038540342e-06,
+      "loss": 0.4637,
+      "step": 1237
+    },
+    {
+      "epoch": 0.12048661800486618,
+      "grad_norm": 1.4918367462943103,
+      "learning_rate": 9.78712561605637e-06,
+      "loss": 0.4998,
+      "step": 1238
+    },
+    {
+      "epoch": 0.12058394160583942,
+      "grad_norm": 1.5775409788682337,
+      "learning_rate": 9.786670371009706e-06,
+      "loss": 0.4415,
+      "step": 1239
+    },
+    {
+      "epoch": 0.12068126520681265,
+      "grad_norm": 1.5427286854632911,
+      "learning_rate": 9.786214650308666e-06,
+      "loss": 0.4606,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12077858880778589,
+      "grad_norm": 1.523821034203494,
+      "learning_rate": 9.78575845399853e-06,
+      "loss": 0.3918,
+      "step": 1241
+    },
+    {
+      "epoch": 0.12087591240875913,
+      "grad_norm": 1.950297391662121,
+      "learning_rate": 9.785301782124638e-06,
+      "loss": 0.5579,
+      "step": 1242
+    },
+    {
+      "epoch": 0.12097323600973237,
+      "grad_norm": 1.5957141815138678,
+      "learning_rate": 9.784844634732367e-06,
+      "loss": 0.3814,
+      "step": 1243
+    },
+    {
+      "epoch": 0.12107055961070559,
+      "grad_norm": 1.3924341327971197,
+      "learning_rate": 9.784387011867145e-06,
+      "loss": 0.3576,
+      "step": 1244
+    },
+    {
+      "epoch": 0.12116788321167883,
+      "grad_norm": 1.670661057733516,
+      "learning_rate": 9.783928913574442e-06,
+      "loss": 0.5307,
+      "step": 1245
+    },
+    {
+      "epoch": 0.12126520681265207,
+      "grad_norm": 1.9162789104592521,
+      "learning_rate": 9.783470339899783e-06,
+      "loss": 0.2309,
+      "step": 1246
+    },
+    {
+      "epoch": 0.12136253041362531,
+      "grad_norm": 1.4323883393925967,
+      "learning_rate": 9.783011290888737e-06,
+      "loss": 0.4816,
+      "step": 1247
+    },
+    {
+      "epoch": 0.12145985401459854,
+      "grad_norm": 1.133557304990043,
+      "learning_rate": 9.78255176658692e-06,
+      "loss": 0.259,
+      "step": 1248
+    },
+    {
+      "epoch": 0.12155717761557178,
+      "grad_norm": 1.6381613262272003,
+      "learning_rate": 9.782091767039992e-06,
+      "loss": 0.535,
+      "step": 1249
+    },
+    {
+      "epoch": 0.12165450121654502,
+      "grad_norm": 1.521879132713644,
+      "learning_rate": 9.781631292293668e-06,
+      "loss": 0.5299,
+      "step": 1250
+    },
+    {
+      "epoch": 0.12175182481751824,
+      "grad_norm": 1.2965362290198492,
+      "learning_rate": 9.781170342393702e-06,
+      "loss": 0.4161,
+      "step": 1251
+    },
+    {
+      "epoch": 0.12184914841849148,
+      "grad_norm": 1.4753461399295356,
+      "learning_rate": 9.780708917385901e-06,
+      "loss": 0.5379,
+      "step": 1252
+    },
+    {
+      "epoch": 0.12194647201946472,
+      "grad_norm": 0.9509628974965367,
+      "learning_rate": 9.780247017316115e-06,
+      "loss": 0.2681,
+      "step": 1253
+    },
+    {
+      "epoch": 0.12204379562043796,
+      "grad_norm": 1.3308735848114122,
+      "learning_rate": 9.779784642230246e-06,
+      "loss": 0.4247,
+      "step": 1254
+    },
+    {
+      "epoch": 0.12214111922141119,
+      "grad_norm": 1.1206835484781008,
+      "learning_rate": 9.779321792174239e-06,
+      "loss": 0.2301,
+      "step": 1255
+    },
+    {
+      "epoch": 0.12223844282238443,
+      "grad_norm": 1.2598096263209464,
+      "learning_rate": 9.778858467194087e-06,
+      "loss": 0.3163,
+      "step": 1256
+    },
+    {
+      "epoch": 0.12233576642335767,
+      "grad_norm": 1.4871998460052394,
+      "learning_rate": 9.778394667335834e-06,
+      "loss": 0.3433,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1224330900243309,
+      "grad_norm": 1.384245738588718,
+      "learning_rate": 9.777930392645565e-06,
+      "loss": 0.2111,
+      "step": 1258
+    },
+    {
+      "epoch": 0.12253041362530413,
+      "grad_norm": 1.4369061113982475,
+      "learning_rate": 9.777465643169417e-06,
+      "loss": 0.3895,
+      "step": 1259
+    },
+    {
+      "epoch": 0.12262773722627737,
+      "grad_norm": 1.8558638944994366,
+      "learning_rate": 9.777000418953568e-06,
+      "loss": 0.3388,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12272506082725061,
+      "grad_norm": 1.512984108492842,
+      "learning_rate": 9.776534720044255e-06,
+      "loss": 0.4726,
+      "step": 1261
+    },
+    {
+      "epoch": 0.12282238442822384,
+      "grad_norm": 1.367540412040702,
+      "learning_rate": 9.77606854648775e-06,
+      "loss": 0.2684,
+      "step": 1262
+    },
+    {
+      "epoch": 0.12291970802919708,
+      "grad_norm": 1.2042550068870583,
+      "learning_rate": 9.775601898330377e-06,
+      "loss": 0.2173,
+      "step": 1263
+    },
+    {
+      "epoch": 0.12301703163017032,
+      "grad_norm": 1.5842484372844456,
+      "learning_rate": 9.775134775618509e-06,
+      "loss": 0.5608,
+      "step": 1264
+    },
+    {
+      "epoch": 0.12311435523114356,
+      "grad_norm": 1.397447971201202,
+      "learning_rate": 9.774667178398562e-06,
+      "loss": 0.4632,
+      "step": 1265
+    },
+    {
+      "epoch": 0.12321167883211678,
+      "grad_norm": 1.3468996882112099,
+      "learning_rate": 9.774199106717004e-06,
+      "loss": 0.3697,
+      "step": 1266
+    },
+    {
+      "epoch": 0.12330900243309002,
+      "grad_norm": 1.252677053550249,
+      "learning_rate": 9.773730560620345e-06,
+      "loss": 0.2377,
+      "step": 1267
+    },
+    {
+      "epoch": 0.12340632603406326,
+      "grad_norm": 1.4179546260918483,
+      "learning_rate": 9.773261540155148e-06,
+      "loss": 0.4857,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1235036496350365,
+      "grad_norm": 1.3092572252570605,
+      "learning_rate": 9.772792045368015e-06,
+      "loss": 0.2969,
+      "step": 1269
+    },
+    {
+      "epoch": 0.12360097323600973,
+      "grad_norm": 1.7901486760202572,
+      "learning_rate": 9.772322076305607e-06,
+      "loss": 0.6935,
+      "step": 1270
+    },
+    {
+      "epoch": 0.12369829683698297,
+      "grad_norm": 1.5982523135009328,
+      "learning_rate": 9.771851633014618e-06,
+      "loss": 0.4368,
+      "step": 1271
+    },
+    {
+      "epoch": 0.12379562043795621,
+      "grad_norm": 1.195950207110724,
+      "learning_rate": 9.7713807155418e-06,
+      "loss": 0.3202,
+      "step": 1272
+    },
+    {
+      "epoch": 0.12389294403892943,
+      "grad_norm": 1.352519407714817,
+      "learning_rate": 9.770909323933947e-06,
+      "loss": 0.4284,
+      "step": 1273
+    },
+    {
+      "epoch": 0.12399026763990267,
+      "grad_norm": 1.4231425912579843,
+      "learning_rate": 9.770437458237903e-06,
+      "loss": 0.434,
+      "step": 1274
+    },
+    {
+      "epoch": 0.12408759124087591,
+      "grad_norm": 1.2825234760121222,
+      "learning_rate": 9.769965118500555e-06,
+      "loss": 0.3817,
+      "step": 1275
+    },
+    {
+      "epoch": 0.12418491484184915,
+      "grad_norm": 1.8250797045299043,
+      "learning_rate": 9.769492304768843e-06,
+      "loss": 0.7366,
+      "step": 1276
+    },
+    {
+      "epoch": 0.12428223844282238,
+      "grad_norm": 1.3974167065714918,
+      "learning_rate": 9.769019017089748e-06,
+      "loss": 0.2804,
+      "step": 1277
+    },
+    {
+      "epoch": 0.12437956204379562,
+      "grad_norm": 1.2933017267383033,
+      "learning_rate": 9.768545255510302e-06,
+      "loss": 0.3495,
+      "step": 1278
+    },
+    {
+      "epoch": 0.12447688564476886,
+      "grad_norm": 1.2423501538003798,
+      "learning_rate": 9.768071020077584e-06,
+      "loss": 0.2908,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1245742092457421,
+      "grad_norm": 1.8228975858143868,
+      "learning_rate": 9.767596310838718e-06,
+      "loss": 0.4222,
+      "step": 1280
+    },
+    {
+      "epoch": 0.12467153284671532,
+      "grad_norm": 1.5510872411682606,
+      "learning_rate": 9.767121127840874e-06,
+      "loss": 0.5058,
+      "step": 1281
+    },
+    {
+      "epoch": 0.12476885644768856,
+      "grad_norm": 1.6665778692750302,
+      "learning_rate": 9.766645471131278e-06,
+      "loss": 0.3592,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1248661800486618,
+      "grad_norm": 1.5396481092124317,
+      "learning_rate": 9.766169340757187e-06,
+      "loss": 0.2737,
+      "step": 1283
+    },
+    {
+      "epoch": 0.12496350364963503,
+      "grad_norm": 1.5555229817491858,
+      "learning_rate": 9.765692736765922e-06,
+      "loss": 0.5466,
+      "step": 1284
+    },
+    {
+      "epoch": 0.12506082725060827,
+      "grad_norm": 1.5351601326386175,
+      "learning_rate": 9.765215659204838e-06,
+      "loss": 0.4733,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1251581508515815,
+      "grad_norm": 1.2793773363741519,
+      "learning_rate": 9.764738108121347e-06,
+      "loss": 0.3056,
+      "step": 1286
+    },
+    {
+      "epoch": 0.12525547445255475,
+      "grad_norm": 1.6331577939793205,
+      "learning_rate": 9.764260083562902e-06,
+      "loss": 0.5883,
+      "step": 1287
+    },
+    {
+      "epoch": 0.12535279805352798,
+      "grad_norm": 1.3363728845544067,
+      "learning_rate": 9.763781585577003e-06,
+      "loss": 0.2904,
+      "step": 1288
+    },
+    {
+      "epoch": 0.12545012165450123,
+      "grad_norm": 1.360818732035961,
+      "learning_rate": 9.763302614211199e-06,
+      "loss": 0.4202,
+      "step": 1289
+    },
+    {
+      "epoch": 0.12554744525547445,
+      "grad_norm": 1.3103877737057137,
+      "learning_rate": 9.762823169513089e-06,
+      "loss": 0.4694,
+      "step": 1290
+    },
+    {
+      "epoch": 0.12564476885644768,
+      "grad_norm": 1.1848446118808063,
+      "learning_rate": 9.76234325153031e-06,
+      "loss": 0.2265,
+      "step": 1291
+    },
+    {
+      "epoch": 0.12574209245742093,
+      "grad_norm": 1.3494947194310234,
+      "learning_rate": 9.761862860310558e-06,
+      "loss": 0.2382,
+      "step": 1292
+    },
+    {
+      "epoch": 0.12583941605839416,
+      "grad_norm": 1.7062717031139596,
+      "learning_rate": 9.761381995901564e-06,
+      "loss": 0.7254,
+      "step": 1293
+    },
+    {
+      "epoch": 0.12593673965936739,
+      "grad_norm": 1.208337515242783,
+      "learning_rate": 9.760900658351117e-06,
+      "loss": 0.326,
+      "step": 1294
+    },
+    {
+      "epoch": 0.12603406326034064,
+      "grad_norm": 1.3159841432369768,
+      "learning_rate": 9.760418847707043e-06,
+      "loss": 0.3438,
+      "step": 1295
+    },
+    {
+      "epoch": 0.12613138686131387,
+      "grad_norm": 1.3809255020300633,
+      "learning_rate": 9.759936564017223e-06,
+      "loss": 0.2716,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1262287104622871,
+      "grad_norm": 1.3382917039666673,
+      "learning_rate": 9.759453807329582e-06,
+      "loss": 0.2882,
+      "step": 1297
+    },
+    {
+      "epoch": 0.12632603406326035,
+      "grad_norm": 1.3572918507167704,
+      "learning_rate": 9.75897057769209e-06,
+      "loss": 0.4181,
+      "step": 1298
+    },
+    {
+      "epoch": 0.12642335766423357,
+      "grad_norm": 1.4433440128897468,
+      "learning_rate": 9.758486875152766e-06,
+      "loss": 0.4883,
+      "step": 1299
+    },
+    {
+      "epoch": 0.12652068126520682,
+      "grad_norm": 1.1934091211117765,
+      "learning_rate": 9.758002699759677e-06,
+      "loss": 0.3828,
+      "step": 1300
+    },
+    {
+      "epoch": 0.12661800486618005,
+      "grad_norm": 1.4647925609545562,
+      "learning_rate": 9.757518051560935e-06,
+      "loss": 0.402,
+      "step": 1301
+    },
+    {
+      "epoch": 0.12671532846715328,
+      "grad_norm": 1.658517832372951,
+      "learning_rate": 9.7570329306047e-06,
+      "loss": 0.6752,
+      "step": 1302
+    },
+    {
+      "epoch": 0.12681265206812653,
+      "grad_norm": 1.2682494280043264,
+      "learning_rate": 9.75654733693918e-06,
+      "loss": 0.2786,
+      "step": 1303
+    },
+    {
+      "epoch": 0.12690997566909976,
+      "grad_norm": 1.3919267883395627,
+      "learning_rate": 9.756061270612625e-06,
+      "loss": 0.4806,
+      "step": 1304
+    },
+    {
+      "epoch": 0.12700729927007298,
+      "grad_norm": 1.160118847382142,
+      "learning_rate": 9.75557473167334e-06,
+      "loss": 0.2458,
+      "step": 1305
+    },
+    {
+      "epoch": 0.12710462287104624,
+      "grad_norm": 1.482640427472728,
+      "learning_rate": 9.755087720169672e-06,
+      "loss": 0.527,
+      "step": 1306
+    },
+    {
+      "epoch": 0.12720194647201946,
+      "grad_norm": 1.5068875178509769,
+      "learning_rate": 9.75460023615001e-06,
+      "loss": 0.4985,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1272992700729927,
+      "grad_norm": 1.2878541774064265,
+      "learning_rate": 9.754112279662805e-06,
+      "loss": 0.3478,
+      "step": 1308
+    },
+    {
+      "epoch": 0.12739659367396594,
+      "grad_norm": 1.1398490461157162,
+      "learning_rate": 9.75362385075654e-06,
+      "loss": 0.3084,
+      "step": 1309
+    },
+    {
+      "epoch": 0.12749391727493917,
+      "grad_norm": 1.2924420070365765,
+      "learning_rate": 9.75313494947975e-06,
+      "loss": 0.3919,
+      "step": 1310
+    },
+    {
+      "epoch": 0.12759124087591242,
+      "grad_norm": 1.4558696462945964,
+      "learning_rate": 9.752645575881018e-06,
+      "loss": 0.225,
+      "step": 1311
+    },
+    {
+      "epoch": 0.12768856447688565,
+      "grad_norm": 1.677251779693783,
+      "learning_rate": 9.752155730008974e-06,
+      "loss": 0.4831,
+      "step": 1312
+    },
+    {
+      "epoch": 0.12778588807785887,
+      "grad_norm": 1.3350720195417478,
+      "learning_rate": 9.751665411912294e-06,
+      "loss": 0.4371,
+      "step": 1313
+    },
+    {
+      "epoch": 0.12788321167883213,
+      "grad_norm": 1.3653039655289896,
+      "learning_rate": 9.751174621639702e-06,
+      "loss": 0.4051,
+      "step": 1314
+    },
+    {
+      "epoch": 0.12798053527980535,
+      "grad_norm": 2.0214110135389927,
+      "learning_rate": 9.75068335923997e-06,
+      "loss": 0.4971,
+      "step": 1315
+    },
+    {
+      "epoch": 0.12807785888077858,
+      "grad_norm": 1.7144522600221743,
+      "learning_rate": 9.750191624761909e-06,
+      "loss": 0.6353,
+      "step": 1316
+    },
+    {
+      "epoch": 0.12817518248175183,
+      "grad_norm": 1.61491787633751,
+      "learning_rate": 9.749699418254388e-06,
+      "loss": 0.5408,
+      "step": 1317
+    },
+    {
+      "epoch": 0.12827250608272506,
+      "grad_norm": 1.3029361322695596,
+      "learning_rate": 9.749206739766317e-06,
+      "loss": 0.407,
+      "step": 1318
+    },
+    {
+      "epoch": 0.12836982968369828,
+      "grad_norm": 1.2453189940624274,
+      "learning_rate": 9.748713589346652e-06,
+      "loss": 0.3254,
+      "step": 1319
+    },
+    {
+      "epoch": 0.12846715328467154,
+      "grad_norm": 1.4117795102544664,
+      "learning_rate": 9.748219967044398e-06,
+      "loss": 0.3941,
+      "step": 1320
+    },
+    {
+      "epoch": 0.12856447688564476,
+      "grad_norm": 1.4197813276706028,
+      "learning_rate": 9.74772587290861e-06,
+      "loss": 0.3454,
+      "step": 1321
+    },
+    {
+      "epoch": 0.12866180048661802,
+      "grad_norm": 1.3133599325252279,
+      "learning_rate": 9.747231306988381e-06,
+      "loss": 0.3389,
+      "step": 1322
+    },
+    {
+      "epoch": 0.12875912408759124,
+      "grad_norm": 1.3432229805022793,
+      "learning_rate": 9.746736269332861e-06,
+      "loss": 0.469,
+      "step": 1323
+    },
+    {
+      "epoch": 0.12885644768856447,
+      "grad_norm": 1.1244292400820686,
+      "learning_rate": 9.746240759991241e-06,
+      "loss": 0.3674,
+      "step": 1324
+    },
+    {
+      "epoch": 0.12895377128953772,
+      "grad_norm": 1.4966792860681473,
+      "learning_rate": 9.745744779012758e-06,
+      "loss": 0.4308,
+      "step": 1325
+    },
+    {
+      "epoch": 0.12905109489051095,
+      "grad_norm": 1.5238028846181695,
+      "learning_rate": 9.745248326446699e-06,
+      "loss": 0.4213,
+      "step": 1326
+    },
+    {
+      "epoch": 0.12914841849148417,
+      "grad_norm": 1.3633303920337936,
+      "learning_rate": 9.744751402342398e-06,
+      "loss": 0.438,
+      "step": 1327
+    },
+    {
+      "epoch": 0.12924574209245743,
+      "grad_norm": 1.3260493495785517,
+      "learning_rate": 9.744254006749235e-06,
+      "loss": 0.4762,
+      "step": 1328
+    },
+    {
+      "epoch": 0.12934306569343065,
+      "grad_norm": 1.705738477220435,
+      "learning_rate": 9.743756139716634e-06,
+      "loss": 0.5861,
+      "step": 1329
+    },
+    {
+      "epoch": 0.12944038929440388,
+      "grad_norm": 1.5829201544013396,
+      "learning_rate": 9.743257801294069e-06,
+      "loss": 0.469,
+      "step": 1330
+    },
+    {
+      "epoch": 0.12953771289537713,
+      "grad_norm": 1.1445128143179795,
+      "learning_rate": 9.74275899153106e-06,
+      "loss": 0.4018,
+      "step": 1331
+    },
+    {
+      "epoch": 0.12963503649635036,
+      "grad_norm": 1.2900129109113572,
+      "learning_rate": 9.742259710477178e-06,
+      "loss": 0.3802,
+      "step": 1332
+    },
+    {
+      "epoch": 0.1297323600973236,
+      "grad_norm": 1.3212461161488713,
+      "learning_rate": 9.74175995818203e-06,
+      "loss": 0.3725,
+      "step": 1333
+    },
+    {
+      "epoch": 0.12982968369829684,
+      "grad_norm": 1.3979706650986563,
+      "learning_rate": 9.741259734695283e-06,
+      "loss": 0.3961,
+      "step": 1334
+    },
+    {
+      "epoch": 0.12992700729927006,
+      "grad_norm": 1.2642819849441118,
+      "learning_rate": 9.740759040066642e-06,
+      "loss": 0.3528,
+      "step": 1335
+    },
+    {
+      "epoch": 0.13002433090024332,
+      "grad_norm": 1.7776493019463793,
+      "learning_rate": 9.74025787434586e-06,
+      "loss": 0.6424,
+      "step": 1336
+    },
+    {
+      "epoch": 0.13012165450121654,
+      "grad_norm": 1.1885806737857232,
+      "learning_rate": 9.73975623758274e-06,
+      "loss": 0.3163,
+      "step": 1337
+    },
+    {
+      "epoch": 0.13021897810218977,
+      "grad_norm": 1.7443954093720497,
+      "learning_rate": 9.739254129827131e-06,
+      "loss": 0.7263,
+      "step": 1338
+    },
+    {
+      "epoch": 0.13031630170316302,
+      "grad_norm": 1.7005058938305366,
+      "learning_rate": 9.738751551128924e-06,
+      "loss": 0.5204,
+      "step": 1339
+    },
+    {
+      "epoch": 0.13041362530413625,
+      "grad_norm": 6.598521165184121,
+      "learning_rate": 9.738248501538063e-06,
+      "loss": 0.5113,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1305109489051095,
+      "grad_norm": 1.6203066466178853,
+      "learning_rate": 9.737744981104536e-06,
+      "loss": 0.625,
+      "step": 1341
+    },
+    {
+      "epoch": 0.13060827250608273,
+      "grad_norm": 1.548111392574701,
+      "learning_rate": 9.73724098987838e-06,
+      "loss": 0.3952,
+      "step": 1342
+    },
+    {
+      "epoch": 0.13070559610705595,
+      "grad_norm": 1.4871418112966692,
+      "learning_rate": 9.736736527909674e-06,
+      "loss": 0.5084,
+      "step": 1343
+    },
+    {
+      "epoch": 0.1308029197080292,
+      "grad_norm": 1.0723677900938815,
+      "learning_rate": 9.736231595248546e-06,
+      "loss": 0.255,
+      "step": 1344
+    },
+    {
+      "epoch": 0.13090024330900243,
+      "grad_norm": 1.5695490713137843,
+      "learning_rate": 9.735726191945176e-06,
+      "loss": 0.3438,
+      "step": 1345
+    },
+    {
+      "epoch": 0.13099756690997566,
+      "grad_norm": 2.059617079542521,
+      "learning_rate": 9.73522031804978e-06,
+      "loss": 0.5249,
+      "step": 1346
+    },
+    {
+      "epoch": 0.1310948905109489,
+      "grad_norm": 1.5301765260275246,
+      "learning_rate": 9.734713973612633e-06,
+      "loss": 0.3667,
+      "step": 1347
+    },
+    {
+      "epoch": 0.13119221411192214,
+      "grad_norm": 1.7431028553023509,
+      "learning_rate": 9.734207158684048e-06,
+      "loss": 0.5551,
+      "step": 1348
+    },
+    {
+      "epoch": 0.13128953771289537,
+      "grad_norm": 1.2916959738739295,
+      "learning_rate": 9.733699873314388e-06,
+      "loss": 0.278,
+      "step": 1349
+    },
+    {
+      "epoch": 0.13138686131386862,
+      "grad_norm": 1.5891072584842363,
+      "learning_rate": 9.733192117554062e-06,
+      "loss": 0.4139,
+      "step": 1350
+    },
+    {
+      "epoch": 0.13148418491484185,
+      "grad_norm": 1.6366778166029219,
+      "learning_rate": 9.732683891453528e-06,
+      "loss": 0.4888,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1315815085158151,
+      "grad_norm": 1.6763551525158185,
+      "learning_rate": 9.732175195063283e-06,
+      "loss": 0.5432,
+      "step": 1352
+    },
+    {
+      "epoch": 0.13167883211678832,
+      "grad_norm": 1.551593865483807,
+      "learning_rate": 9.731666028433882e-06,
+      "loss": 0.5634,
+      "step": 1353
+    },
+    {
+      "epoch": 0.13177615571776155,
+      "grad_norm": 1.693219206573502,
+      "learning_rate": 9.731156391615919e-06,
+      "loss": 0.4554,
+      "step": 1354
+    },
+    {
+      "epoch": 0.1318734793187348,
+      "grad_norm": 1.4894832853139421,
+      "learning_rate": 9.730646284660037e-06,
+      "loss": 0.4286,
+      "step": 1355
+    },
+    {
+      "epoch": 0.13197080291970803,
+      "grad_norm": 1.20058966692396,
+      "learning_rate": 9.730135707616927e-06,
+      "loss": 0.2519,
+      "step": 1356
+    },
+    {
+      "epoch": 0.13206812652068126,
+      "grad_norm": 1.395115321325138,
+      "learning_rate": 9.729624660537324e-06,
+      "loss": 0.3718,
+      "step": 1357
+    },
+    {
+      "epoch": 0.1321654501216545,
+      "grad_norm": 1.3441869335850034,
+      "learning_rate": 9.729113143472011e-06,
+      "loss": 0.43,
+      "step": 1358
+    },
+    {
+      "epoch": 0.13226277372262774,
+      "grad_norm": 1.31865416445236,
+      "learning_rate": 9.72860115647182e-06,
+      "loss": 0.296,
+      "step": 1359
+    },
+    {
+      "epoch": 0.13236009732360096,
+      "grad_norm": 1.3998148863889133,
+      "learning_rate": 9.728088699587623e-06,
+      "loss": 0.2642,
+      "step": 1360
+    },
+    {
+      "epoch": 0.13245742092457422,
+      "grad_norm": 1.5917388343760925,
+      "learning_rate": 9.727575772870347e-06,
+      "loss": 0.5999,
+      "step": 1361
+    },
+    {
+      "epoch": 0.13255474452554744,
+      "grad_norm": 1.6062441992747731,
+      "learning_rate": 9.727062376370962e-06,
+      "loss": 0.6017,
+      "step": 1362
+    },
+    {
+      "epoch": 0.1326520681265207,
+      "grad_norm": 1.756325054261889,
+      "learning_rate": 9.72654851014048e-06,
+      "loss": 0.5855,
+      "step": 1363
+    },
+    {
+      "epoch": 0.13274939172749392,
+      "grad_norm": 1.5782112626775713,
+      "learning_rate": 9.72603417422997e-06,
+      "loss": 0.5643,
+      "step": 1364
+    },
+    {
+      "epoch": 0.13284671532846715,
+      "grad_norm": 1.6280008631148617,
+      "learning_rate": 9.725519368690539e-06,
+      "loss": 0.3918,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1329440389294404,
+      "grad_norm": 1.731476294535625,
+      "learning_rate": 9.725004093573343e-06,
+      "loss": 0.6909,
+      "step": 1366
+    },
+    {
+      "epoch": 0.13304136253041363,
+      "grad_norm": 1.7012591859680217,
+      "learning_rate": 9.724488348929587e-06,
+      "loss": 0.3206,
+      "step": 1367
+    },
+    {
+      "epoch": 0.13313868613138685,
+      "grad_norm": 1.5539166250213363,
+      "learning_rate": 9.723972134810519e-06,
+      "loss": 0.3735,
+      "step": 1368
+    },
+    {
+      "epoch": 0.1332360097323601,
+      "grad_norm": 1.2431527472113675,
+      "learning_rate": 9.723455451267436e-06,
+      "loss": 0.4023,
+      "step": 1369
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 1.4147928785913308,
+      "learning_rate": 9.722938298351682e-06,
+      "loss": 0.4501,
+      "step": 1370
+    },
+    {
+      "epoch": 0.13343065693430656,
+      "grad_norm": 1.5209109752466956,
+      "learning_rate": 9.722420676114646e-06,
+      "loss": 0.4504,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1335279805352798,
+      "grad_norm": 1.6031977794999224,
+      "learning_rate": 9.721902584607766e-06,
+      "loss": 0.4036,
+      "step": 1372
+    },
+    {
+      "epoch": 0.13362530413625304,
+      "grad_norm": 1.3752266957066934,
+      "learning_rate": 9.721384023882524e-06,
+      "loss": 0.4008,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1337226277372263,
+      "grad_norm": 1.2289957585024915,
+      "learning_rate": 9.720864993990448e-06,
+      "loss": 0.3214,
+      "step": 1374
+    },
+    {
+      "epoch": 0.13381995133819952,
+      "grad_norm": 1.5334770200964671,
+      "learning_rate": 9.720345494983117e-06,
+      "loss": 0.4103,
+      "step": 1375
+    },
+    {
+      "epoch": 0.13391727493917274,
+      "grad_norm": 1.4428318489533865,
+      "learning_rate": 9.719825526912152e-06,
+      "loss": 0.4314,
+      "step": 1376
+    },
+    {
+      "epoch": 0.134014598540146,
+      "grad_norm": 1.6794168653476527,
+      "learning_rate": 9.719305089829224e-06,
+      "loss": 0.6027,
+      "step": 1377
+    },
+    {
+      "epoch": 0.13411192214111922,
+      "grad_norm": 1.4695816931820398,
+      "learning_rate": 9.718784183786048e-06,
+      "loss": 0.5337,
+      "step": 1378
+    },
+    {
+      "epoch": 0.13420924574209245,
+      "grad_norm": 1.428180254445363,
+      "learning_rate": 9.718262808834386e-06,
+      "loss": 0.3636,
+      "step": 1379
+    },
+    {
+      "epoch": 0.1343065693430657,
+      "grad_norm": 1.4446624118640763,
+      "learning_rate": 9.717740965026051e-06,
+      "loss": 0.4213,
+      "step": 1380
+    },
+    {
+      "epoch": 0.13440389294403893,
+      "grad_norm": 1.0145899854020284,
+      "learning_rate": 9.717218652412896e-06,
+      "loss": 0.292,
+      "step": 1381
+    },
+    {
+      "epoch": 0.13450121654501215,
+      "grad_norm": 1.4589445831305994,
+      "learning_rate": 9.716695871046824e-06,
+      "loss": 0.4787,
+      "step": 1382
+    },
+    {
+      "epoch": 0.1345985401459854,
+      "grad_norm": 1.5462880417778382,
+      "learning_rate": 9.716172620979783e-06,
+      "loss": 0.4716,
+      "step": 1383
+    },
+    {
+      "epoch": 0.13469586374695863,
+      "grad_norm": 1.5411526965931406,
+      "learning_rate": 9.71564890226377e-06,
+      "loss": 0.5311,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1347931873479319,
+      "grad_norm": 1.1928924795974905,
+      "learning_rate": 9.71512471495083e-06,
+      "loss": 0.2724,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1348905109489051,
+      "grad_norm": 1.3201585939530793,
+      "learning_rate": 9.714600059093045e-06,
+      "loss": 0.2987,
+      "step": 1386
+    },
+    {
+      "epoch": 0.13498783454987834,
+      "grad_norm": 1.5687746327202647,
+      "learning_rate": 9.714074934742556e-06,
+      "loss": 0.363,
+      "step": 1387
+    },
+    {
+      "epoch": 0.1350851581508516,
+      "grad_norm": 1.5338932500845779,
+      "learning_rate": 9.713549341951543e-06,
+      "loss": 0.5661,
+      "step": 1388
+    },
+    {
+      "epoch": 0.13518248175182482,
+      "grad_norm": 1.1601153536694444,
+      "learning_rate": 9.713023280772236e-06,
+      "loss": 0.3079,
+      "step": 1389
+    },
+    {
+      "epoch": 0.13527980535279804,
+      "grad_norm": 1.3983637614495477,
+      "learning_rate": 9.712496751256907e-06,
+      "loss": 0.4741,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1353771289537713,
+      "grad_norm": 1.2378967843544995,
+      "learning_rate": 9.71196975345788e-06,
+      "loss": 0.3467,
+      "step": 1391
+    },
+    {
+      "epoch": 0.13547445255474452,
+      "grad_norm": 1.3128641622430697,
+      "learning_rate": 9.711442287427523e-06,
+      "loss": 0.413,
+      "step": 1392
+    },
+    {
+      "epoch": 0.13557177615571775,
+      "grad_norm": 1.6638151172989781,
+      "learning_rate": 9.71091435321825e-06,
+      "loss": 0.4844,
+      "step": 1393
+    },
+    {
+      "epoch": 0.135669099756691,
+      "grad_norm": 1.5023430961651105,
+      "learning_rate": 9.710385950882522e-06,
+      "loss": 0.3639,
+      "step": 1394
+    },
+    {
+      "epoch": 0.13576642335766423,
+      "grad_norm": 1.3286069107884302,
+      "learning_rate": 9.709857080472847e-06,
+      "loss": 0.4055,
+      "step": 1395
+    },
+    {
+      "epoch": 0.13586374695863748,
+      "grad_norm": 1.2934746343236392,
+      "learning_rate": 9.709327742041776e-06,
+      "loss": 0.2837,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1359610705596107,
+      "grad_norm": 1.698077360010743,
+      "learning_rate": 9.708797935641915e-06,
+      "loss": 0.3687,
+      "step": 1397
+    },
+    {
+      "epoch": 0.13605839416058393,
+      "grad_norm": 1.412271661785088,
+      "learning_rate": 9.70826766132591e-06,
+      "loss": 0.3577,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1361557177615572,
+      "grad_norm": 1.5421055950766074,
+      "learning_rate": 9.707736919146453e-06,
+      "loss": 0.5394,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1362530413625304,
+      "grad_norm": 2.554386599490806,
+      "learning_rate": 9.707205709156285e-06,
+      "loss": 0.212,
+      "step": 1400
+    },
+    {
+      "epoch": 0.13635036496350364,
+      "grad_norm": 1.7436805109650844,
+      "learning_rate": 9.70667403140819e-06,
+      "loss": 0.6893,
+      "step": 1401
+    },
+    {
+      "epoch": 0.1364476885644769,
+      "grad_norm": 1.613527884115612,
+      "learning_rate": 9.706141885955006e-06,
+      "loss": 0.42,
+      "step": 1402
+    },
+    {
+      "epoch": 0.13654501216545012,
+      "grad_norm": 1.711619341430359,
+      "learning_rate": 9.70560927284961e-06,
+      "loss": 0.7025,
+      "step": 1403
+    },
+    {
+      "epoch": 0.13664233576642335,
+      "grad_norm": 1.5376532439489434,
+      "learning_rate": 9.705076192144927e-06,
+      "loss": 0.5201,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1367396593673966,
+      "grad_norm": 1.492510855426001,
+      "learning_rate": 9.704542643893931e-06,
+      "loss": 0.4281,
+      "step": 1405
+    },
+    {
+      "epoch": 0.13683698296836982,
+      "grad_norm": 1.5678573317920237,
+      "learning_rate": 9.704008628149641e-06,
+      "loss": 0.506,
+      "step": 1406
+    },
+    {
+      "epoch": 0.13693430656934308,
+      "grad_norm": 1.3237691920747017,
+      "learning_rate": 9.703474144965123e-06,
+      "loss": 0.4114,
+      "step": 1407
+    },
+    {
+      "epoch": 0.1370316301703163,
+      "grad_norm": 1.4134135574988251,
+      "learning_rate": 9.702939194393489e-06,
+      "loss": 0.3806,
+      "step": 1408
+    },
+    {
+      "epoch": 0.13712895377128953,
+      "grad_norm": 1.5544258549266206,
+      "learning_rate": 9.702403776487895e-06,
+      "loss": 0.4863,
+      "step": 1409
+    },
+    {
+      "epoch": 0.13722627737226278,
+      "grad_norm": 1.3619063912879554,
+      "learning_rate": 9.701867891301548e-06,
+      "loss": 0.3692,
+      "step": 1410
+    },
+    {
+      "epoch": 0.137323600973236,
+      "grad_norm": 1.5146665393724075,
+      "learning_rate": 9.701331538887699e-06,
+      "loss": 0.3311,
+      "step": 1411
+    },
+    {
+      "epoch": 0.13742092457420924,
+      "grad_norm": 1.5674647990142176,
+      "learning_rate": 9.700794719299644e-06,
+      "loss": 0.5292,
+      "step": 1412
+    },
+    {
+      "epoch": 0.1375182481751825,
+      "grad_norm": 1.4711236643775818,
+      "learning_rate": 9.700257432590729e-06,
+      "loss": 0.466,
+      "step": 1413
+    },
+    {
+      "epoch": 0.13761557177615572,
+      "grad_norm": 1.4410106250389758,
+      "learning_rate": 9.699719678814345e-06,
+      "loss": 0.3276,
+      "step": 1414
+    },
+    {
+      "epoch": 0.13771289537712894,
+      "grad_norm": 1.652937978394441,
+      "learning_rate": 9.699181458023927e-06,
+      "loss": 0.5391,
+      "step": 1415
+    },
+    {
+      "epoch": 0.1378102189781022,
+      "grad_norm": 1.7285587973510355,
+      "learning_rate": 9.698642770272959e-06,
+      "loss": 0.5707,
+      "step": 1416
+    },
+    {
+      "epoch": 0.13790754257907542,
+      "grad_norm": 1.325058257423692,
+      "learning_rate": 9.698103615614972e-06,
+      "loss": 0.3429,
+      "step": 1417
+    },
+    {
+      "epoch": 0.13800486618004867,
+      "grad_norm": 1.5653351048996198,
+      "learning_rate": 9.69756399410354e-06,
+      "loss": 0.4132,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1381021897810219,
+      "grad_norm": 1.603805088396393,
+      "learning_rate": 9.697023905792287e-06,
+      "loss": 0.4983,
+      "step": 1419
+    },
+    {
+      "epoch": 0.13819951338199513,
+      "grad_norm": 1.5052443063346659,
+      "learning_rate": 9.69648335073488e-06,
+      "loss": 0.2713,
+      "step": 1420
+    },
+    {
+      "epoch": 0.13829683698296838,
+      "grad_norm": 1.30196768692164,
+      "learning_rate": 9.695942328985037e-06,
+      "loss": 0.27,
+      "step": 1421
+    },
+    {
+      "epoch": 0.1383941605839416,
+      "grad_norm": 1.1542739478608208,
+      "learning_rate": 9.695400840596519e-06,
+      "loss": 0.3309,
+      "step": 1422
+    },
+    {
+      "epoch": 0.13849148418491483,
+      "grad_norm": 1.1029138054910885,
+      "learning_rate": 9.694858885623132e-06,
+      "loss": 0.3262,
+      "step": 1423
+    },
+    {
+      "epoch": 0.13858880778588809,
+      "grad_norm": 1.581389120261872,
+      "learning_rate": 9.694316464118732e-06,
+      "loss": 0.4663,
+      "step": 1424
+    },
+    {
+      "epoch": 0.1386861313868613,
+      "grad_norm": 1.2966198038166061,
+      "learning_rate": 9.69377357613722e-06,
+      "loss": 0.336,
+      "step": 1425
+    },
+    {
+      "epoch": 0.13878345498783454,
+      "grad_norm": 1.505634533514273,
+      "learning_rate": 9.693230221732544e-06,
+      "loss": 0.4269,
+      "step": 1426
+    },
+    {
+      "epoch": 0.1388807785888078,
+      "grad_norm": 1.274453115047599,
+      "learning_rate": 9.692686400958695e-06,
+      "loss": 0.3978,
+      "step": 1427
+    },
+    {
+      "epoch": 0.13897810218978102,
+      "grad_norm": 1.2126154933077449,
+      "learning_rate": 9.692142113869714e-06,
+      "loss": 0.2754,
+      "step": 1428
+    },
+    {
+      "epoch": 0.13907542579075427,
+      "grad_norm": 1.4884313472642259,
+      "learning_rate": 9.691597360519686e-06,
+      "loss": 0.4661,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1391727493917275,
+      "grad_norm": 1.5680101511782372,
+      "learning_rate": 9.691052140962747e-06,
+      "loss": 0.4237,
+      "step": 1430
+    },
+    {
+      "epoch": 0.13927007299270072,
+      "grad_norm": 1.325640699841282,
+      "learning_rate": 9.690506455253073e-06,
+      "loss": 0.3988,
+      "step": 1431
+    },
+    {
+      "epoch": 0.13936739659367398,
+      "grad_norm": 1.3107002270910884,
+      "learning_rate": 9.689960303444887e-06,
+      "loss": 0.4268,
+      "step": 1432
+    },
+    {
+      "epoch": 0.1394647201946472,
+      "grad_norm": 1.9246823036308274,
+      "learning_rate": 9.689413685592465e-06,
+      "loss": 0.3733,
+      "step": 1433
+    },
+    {
+      "epoch": 0.13956204379562043,
+      "grad_norm": 1.3731854343094059,
+      "learning_rate": 9.688866601750122e-06,
+      "loss": 0.4215,
+      "step": 1434
+    },
+    {
+      "epoch": 0.13965936739659368,
+      "grad_norm": 1.368964734934982,
+      "learning_rate": 9.688319051972224e-06,
+      "loss": 0.4697,
+      "step": 1435
+    },
+    {
+      "epoch": 0.1397566909975669,
+      "grad_norm": 1.3451140821212522,
+      "learning_rate": 9.687771036313178e-06,
+      "loss": 0.3741,
+      "step": 1436
+    },
+    {
+      "epoch": 0.13985401459854013,
+      "grad_norm": 1.5372748667563303,
+      "learning_rate": 9.687222554827444e-06,
+      "loss": 0.4199,
+      "step": 1437
+    },
+    {
+      "epoch": 0.1399513381995134,
+      "grad_norm": 1.1780522614950486,
+      "learning_rate": 9.686673607569526e-06,
+      "loss": 0.3602,
+      "step": 1438
+    },
+    {
+      "epoch": 0.1400486618004866,
+      "grad_norm": 1.20778383169165,
+      "learning_rate": 9.686124194593967e-06,
+      "loss": 0.23,
+      "step": 1439
+    },
+    {
+      "epoch": 0.14014598540145987,
+      "grad_norm": 1.6760972087501165,
+      "learning_rate": 9.685574315955368e-06,
+      "loss": 0.5089,
+      "step": 1440
+    },
+    {
+      "epoch": 0.1402433090024331,
+      "grad_norm": 1.7963497555189056,
+      "learning_rate": 9.68502397170837e-06,
+      "loss": 0.3932,
+      "step": 1441
+    },
+    {
+      "epoch": 0.14034063260340632,
+      "grad_norm": 1.401968265514402,
+      "learning_rate": 9.68447316190766e-06,
+      "loss": 0.4272,
+      "step": 1442
+    },
+    {
+      "epoch": 0.14043795620437957,
+      "grad_norm": 1.1461895591250986,
+      "learning_rate": 9.683921886607973e-06,
+      "loss": 0.3003,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1405352798053528,
+      "grad_norm": 1.8257595963636586,
+      "learning_rate": 9.683370145864089e-06,
+      "loss": 0.4454,
+      "step": 1444
+    },
+    {
+      "epoch": 0.14063260340632602,
+      "grad_norm": 1.3483599166387192,
+      "learning_rate": 9.682817939730833e-06,
+      "loss": 0.3708,
+      "step": 1445
+    },
+    {
+      "epoch": 0.14072992700729928,
+      "grad_norm": 1.4560700792487955,
+      "learning_rate": 9.682265268263083e-06,
+      "loss": 0.4321,
+      "step": 1446
+    },
+    {
+      "epoch": 0.1408272506082725,
+      "grad_norm": 1.4364952224667933,
+      "learning_rate": 9.681712131515753e-06,
+      "loss": 0.3812,
+      "step": 1447
+    },
+    {
+      "epoch": 0.14092457420924573,
+      "grad_norm": 1.6808986821455574,
+      "learning_rate": 9.681158529543812e-06,
+      "loss": 0.3939,
+      "step": 1448
+    },
+    {
+      "epoch": 0.14102189781021898,
+      "grad_norm": 1.5327313322922438,
+      "learning_rate": 9.68060446240227e-06,
+      "loss": 0.3617,
+      "step": 1449
+    },
+    {
+      "epoch": 0.1411192214111922,
+      "grad_norm": 1.9055650449775412,
+      "learning_rate": 9.680049930146186e-06,
+      "loss": 0.4984,
+      "step": 1450
+    },
+    {
+      "epoch": 0.14121654501216546,
+      "grad_norm": 1.8971706606162055,
+      "learning_rate": 9.679494932830664e-06,
+      "loss": 0.4196,
+      "step": 1451
+    },
+    {
+      "epoch": 0.1413138686131387,
+      "grad_norm": 1.7337796675846617,
+      "learning_rate": 9.678939470510856e-06,
+      "loss": 0.4282,
+      "step": 1452
+    },
+    {
+      "epoch": 0.1414111922141119,
+      "grad_norm": 1.6436762455975924,
+      "learning_rate": 9.678383543241954e-06,
+      "loss": 0.425,
+      "step": 1453
+    },
+    {
+      "epoch": 0.14150851581508517,
+      "grad_norm": 1.3304471527694197,
+      "learning_rate": 9.677827151079205e-06,
+      "loss": 0.346,
+      "step": 1454
+    },
+    {
+      "epoch": 0.1416058394160584,
+      "grad_norm": 1.3162532004293022,
+      "learning_rate": 9.677270294077895e-06,
+      "loss": 0.4492,
+      "step": 1455
+    },
+    {
+      "epoch": 0.14170316301703162,
+      "grad_norm": 1.2299075830057253,
+      "learning_rate": 9.676712972293363e-06,
+      "loss": 0.3525,
+      "step": 1456
+    },
+    {
+      "epoch": 0.14180048661800487,
+      "grad_norm": 1.7174455721253266,
+      "learning_rate": 9.676155185780989e-06,
+      "loss": 0.763,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1418978102189781,
+      "grad_norm": 0.9624475539149472,
+      "learning_rate": 9.675596934596198e-06,
+      "loss": 0.2234,
+      "step": 1458
+    },
+    {
+      "epoch": 0.14199513381995132,
+      "grad_norm": 1.380722751360302,
+      "learning_rate": 9.675038218794469e-06,
+      "loss": 0.3539,
+      "step": 1459
+    },
+    {
+      "epoch": 0.14209245742092458,
+      "grad_norm": 1.3595616004290971,
+      "learning_rate": 9.674479038431314e-06,
+      "loss": 0.4356,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1421897810218978,
+      "grad_norm": 1.2777542247997187,
+      "learning_rate": 9.673919393562308e-06,
+      "loss": 0.3233,
+      "step": 1461
+    },
+    {
+      "epoch": 0.14228710462287106,
+      "grad_norm": 1.23752096524445,
+      "learning_rate": 9.673359284243055e-06,
+      "loss": 0.405,
+      "step": 1462
+    },
+    {
+      "epoch": 0.14238442822384428,
+      "grad_norm": 1.4547729172095425,
+      "learning_rate": 9.672798710529222e-06,
+      "loss": 0.5356,
+      "step": 1463
+    },
+    {
+      "epoch": 0.1424817518248175,
+      "grad_norm": 1.5976011084855026,
+      "learning_rate": 9.672237672476506e-06,
+      "loss": 0.571,
+      "step": 1464
+    },
+    {
+      "epoch": 0.14257907542579076,
+      "grad_norm": 1.4454139467669962,
+      "learning_rate": 9.67167617014066e-06,
+      "loss": 0.4556,
+      "step": 1465
+    },
+    {
+      "epoch": 0.142676399026764,
+      "grad_norm": 1.5296734849172828,
+      "learning_rate": 9.671114203577485e-06,
+      "loss": 0.5791,
+      "step": 1466
+    },
+    {
+      "epoch": 0.14277372262773722,
+      "grad_norm": 1.0140913901893902,
+      "learning_rate": 9.670551772842818e-06,
+      "loss": 0.2732,
+      "step": 1467
+    },
+    {
+      "epoch": 0.14287104622871047,
+      "grad_norm": 1.5600773149062541,
+      "learning_rate": 9.669988877992551e-06,
+      "loss": 0.3902,
+      "step": 1468
+    },
+    {
+      "epoch": 0.1429683698296837,
+      "grad_norm": 1.4767158872669255,
+      "learning_rate": 9.66942551908262e-06,
+      "loss": 0.5531,
+      "step": 1469
+    },
+    {
+      "epoch": 0.14306569343065692,
+      "grad_norm": 1.1134570066684917,
+      "learning_rate": 9.668861696169003e-06,
+      "loss": 0.278,
+      "step": 1470
+    },
+    {
+      "epoch": 0.14316301703163017,
+      "grad_norm": 0.9776488708344422,
+      "learning_rate": 9.66829740930773e-06,
+      "loss": 0.231,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1432603406326034,
+      "grad_norm": 1.4647496714581032,
+      "learning_rate": 9.667732658554875e-06,
+      "loss": 0.485,
+      "step": 1472
+    },
+    {
+      "epoch": 0.14335766423357665,
+      "grad_norm": 1.2234301570511203,
+      "learning_rate": 9.667167443966557e-06,
+      "loss": 0.3944,
+      "step": 1473
+    },
+    {
+      "epoch": 0.14345498783454988,
+      "grad_norm": 1.3655487702696618,
+      "learning_rate": 9.66660176559894e-06,
+      "loss": 0.3989,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1435523114355231,
+      "grad_norm": 1.4690108372007447,
+      "learning_rate": 9.666035623508238e-06,
+      "loss": 0.4311,
+      "step": 1475
+    },
+    {
+      "epoch": 0.14364963503649636,
+      "grad_norm": 1.1910374305057687,
+      "learning_rate": 9.665469017750708e-06,
+      "loss": 0.3002,
+      "step": 1476
+    },
+    {
+      "epoch": 0.14374695863746959,
+      "grad_norm": 1.678176413091249,
+      "learning_rate": 9.664901948382654e-06,
+      "loss": 0.6143,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1438442822384428,
+      "grad_norm": 1.817046546881487,
+      "learning_rate": 9.664334415460426e-06,
+      "loss": 0.7811,
+      "step": 1478
+    },
+    {
+      "epoch": 0.14394160583941606,
+      "grad_norm": 1.4955026439922687,
+      "learning_rate": 9.663766419040422e-06,
+      "loss": 0.411,
+      "step": 1479
+    },
+    {
+      "epoch": 0.1440389294403893,
+      "grad_norm": 1.4198677231066263,
+      "learning_rate": 9.66319795917908e-06,
+      "loss": 0.4245,
+      "step": 1480
+    },
+    {
+      "epoch": 0.14413625304136254,
+      "grad_norm": 1.5199876898969789,
+      "learning_rate": 9.662629035932892e-06,
+      "loss": 0.438,
+      "step": 1481
+    },
+    {
+      "epoch": 0.14423357664233577,
+      "grad_norm": 1.4859771113526168,
+      "learning_rate": 9.662059649358388e-06,
+      "loss": 0.3949,
+      "step": 1482
+    },
+    {
+      "epoch": 0.144330900243309,
+      "grad_norm": 1.5386966328542977,
+      "learning_rate": 9.661489799512155e-06,
+      "loss": 0.4679,
+      "step": 1483
+    },
+    {
+      "epoch": 0.14442822384428225,
+      "grad_norm": 1.2872766782537612,
+      "learning_rate": 9.660919486450813e-06,
+      "loss": 0.2624,
+      "step": 1484
+    },
+    {
+      "epoch": 0.14452554744525548,
+      "grad_norm": 1.3276179523832277,
+      "learning_rate": 9.660348710231037e-06,
+      "loss": 0.5476,
+      "step": 1485
+    },
+    {
+      "epoch": 0.1446228710462287,
+      "grad_norm": 0.9490583621811937,
+      "learning_rate": 9.659777470909547e-06,
+      "loss": 0.2354,
+      "step": 1486
+    },
+    {
+      "epoch": 0.14472019464720196,
+      "grad_norm": 1.3763558898436123,
+      "learning_rate": 9.659205768543104e-06,
+      "loss": 0.4327,
+      "step": 1487
+    },
+    {
+      "epoch": 0.14481751824817518,
+      "grad_norm": 1.178366926128956,
+      "learning_rate": 9.658633603188521e-06,
+      "loss": 0.3839,
+      "step": 1488
+    },
+    {
+      "epoch": 0.1449148418491484,
+      "grad_norm": 1.3255542333725456,
+      "learning_rate": 9.658060974902653e-06,
+      "loss": 0.3068,
+      "step": 1489
+    },
+    {
+      "epoch": 0.14501216545012166,
+      "grad_norm": 1.5998345706772108,
+      "learning_rate": 9.657487883742403e-06,
+      "loss": 0.5432,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1451094890510949,
+      "grad_norm": 1.8804975658787435,
+      "learning_rate": 9.656914329764718e-06,
+      "loss": 0.5268,
+      "step": 1491
+    },
+    {
+      "epoch": 0.14520681265206814,
+      "grad_norm": 1.5841269093835124,
+      "learning_rate": 9.656340313026595e-06,
+      "loss": 0.6304,
+      "step": 1492
+    },
+    {
+      "epoch": 0.14530413625304137,
+      "grad_norm": 1.5832299483159056,
+      "learning_rate": 9.655765833585072e-06,
+      "loss": 0.4417,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1454014598540146,
+      "grad_norm": 1.2541361090062475,
+      "learning_rate": 9.655190891497237e-06,
+      "loss": 0.2956,
+      "step": 1494
+    },
+    {
+      "epoch": 0.14549878345498785,
+      "grad_norm": 1.4549578520972333,
+      "learning_rate": 9.654615486820223e-06,
+      "loss": 0.5352,
+      "step": 1495
+    },
+    {
+      "epoch": 0.14559610705596107,
+      "grad_norm": 1.4797996277102474,
+      "learning_rate": 9.654039619611205e-06,
+      "loss": 0.4915,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1456934306569343,
+      "grad_norm": 1.2281886698207842,
+      "learning_rate": 9.65346328992741e-06,
+      "loss": 0.1901,
+      "step": 1497
+    },
+    {
+      "epoch": 0.14579075425790755,
+      "grad_norm": 1.4478972545758728,
+      "learning_rate": 9.652886497826109e-06,
+      "loss": 0.4142,
+      "step": 1498
+    },
+    {
+      "epoch": 0.14588807785888078,
+      "grad_norm": 1.5883286963945868,
+      "learning_rate": 9.652309243364614e-06,
+      "loss": 0.3576,
+      "step": 1499
+    },
+    {
+      "epoch": 0.145985401459854,
+      "grad_norm": 1.5369489845441549,
+      "learning_rate": 9.651731526600293e-06,
+      "loss": 0.5479,
+      "step": 1500
+    },
+    {
+      "epoch": 0.14608272506082726,
+      "grad_norm": 1.5655077404950533,
+      "learning_rate": 9.651153347590549e-06,
+      "loss": 0.3464,
+      "step": 1501
+    },
+    {
+      "epoch": 0.14618004866180048,
+      "grad_norm": 1.6426065013852054,
+      "learning_rate": 9.65057470639284e-06,
+      "loss": 0.5038,
+      "step": 1502
+    },
+    {
+      "epoch": 0.14627737226277374,
+      "grad_norm": 1.8088684532537898,
+      "learning_rate": 9.649995603064664e-06,
+      "loss": 0.5731,
+      "step": 1503
+    },
+    {
+      "epoch": 0.14637469586374696,
+      "grad_norm": 1.2493389766016731,
+      "learning_rate": 9.649416037663564e-06,
+      "loss": 0.3306,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1464720194647202,
+      "grad_norm": 1.5964615139072293,
+      "learning_rate": 9.648836010247137e-06,
+      "loss": 0.4169,
+      "step": 1505
+    },
+    {
+      "epoch": 0.14656934306569344,
+      "grad_norm": 1.3925830899828215,
+      "learning_rate": 9.648255520873018e-06,
+      "loss": 0.3092,
+      "step": 1506
+    },
+    {
+      "epoch": 0.14666666666666667,
+      "grad_norm": 1.4256860988690832,
+      "learning_rate": 9.647674569598889e-06,
+      "loss": 0.3201,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1467639902676399,
+      "grad_norm": 1.6614703553660697,
+      "learning_rate": 9.647093156482483e-06,
+      "loss": 0.6078,
+      "step": 1508
+    },
+    {
+      "epoch": 0.14686131386861315,
+      "grad_norm": 1.4357097092225446,
+      "learning_rate": 9.646511281581575e-06,
+      "loss": 0.4004,
+      "step": 1509
+    },
+    {
+      "epoch": 0.14695863746958637,
+      "grad_norm": 1.4562846462074024,
+      "learning_rate": 9.645928944953981e-06,
+      "loss": 0.4601,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1470559610705596,
+      "grad_norm": 1.3277232740610976,
+      "learning_rate": 9.645346146657575e-06,
+      "loss": 0.4015,
+      "step": 1511
+    },
+    {
+      "epoch": 0.14715328467153285,
+      "grad_norm": 1.5964514332978577,
+      "learning_rate": 9.644762886750267e-06,
+      "loss": 0.4556,
+      "step": 1512
+    },
+    {
+      "epoch": 0.14725060827250608,
+      "grad_norm": 1.4663379423625913,
+      "learning_rate": 9.644179165290015e-06,
+      "loss": 0.4353,
+      "step": 1513
+    },
+    {
+      "epoch": 0.14734793187347933,
+      "grad_norm": 1.0949765548634744,
+      "learning_rate": 9.643594982334826e-06,
+      "loss": 0.2276,
+      "step": 1514
+    },
+    {
+      "epoch": 0.14744525547445256,
+      "grad_norm": 1.563845779693575,
+      "learning_rate": 9.643010337942749e-06,
+      "loss": 0.6694,
+      "step": 1515
+    },
+    {
+      "epoch": 0.14754257907542578,
+      "grad_norm": 1.024413538842663,
+      "learning_rate": 9.642425232171881e-06,
+      "loss": 0.3047,
+      "step": 1516
+    },
+    {
+      "epoch": 0.14763990267639904,
+      "grad_norm": 1.712866405633365,
+      "learning_rate": 9.641839665080363e-06,
+      "loss": 0.6729,
+      "step": 1517
+    },
+    {
+      "epoch": 0.14773722627737226,
+      "grad_norm": 1.4526258041869373,
+      "learning_rate": 9.641253636726386e-06,
+      "loss": 0.5037,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1478345498783455,
+      "grad_norm": 1.7375410582816389,
+      "learning_rate": 9.640667147168182e-06,
+      "loss": 0.6717,
+      "step": 1519
+    },
+    {
+      "epoch": 0.14793187347931874,
+      "grad_norm": 1.736227335112512,
+      "learning_rate": 9.640080196464032e-06,
+      "loss": 0.6677,
+      "step": 1520
+    },
+    {
+      "epoch": 0.14802919708029197,
+      "grad_norm": 1.5194007013329096,
+      "learning_rate": 9.63949278467226e-06,
+      "loss": 0.4288,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1481265206812652,
+      "grad_norm": 1.5063763039212688,
+      "learning_rate": 9.638904911851237e-06,
+      "loss": 0.4529,
+      "step": 1522
+    },
+    {
+      "epoch": 0.14822384428223845,
+      "grad_norm": 1.6414876214129155,
+      "learning_rate": 9.638316578059384e-06,
+      "loss": 0.5482,
+      "step": 1523
+    },
+    {
+      "epoch": 0.14832116788321167,
+      "grad_norm": 1.3122113228270877,
+      "learning_rate": 9.63772778335516e-06,
+      "loss": 0.3903,
+      "step": 1524
+    },
+    {
+      "epoch": 0.14841849148418493,
+      "grad_norm": 1.6417051120393822,
+      "learning_rate": 9.637138527797075e-06,
+      "loss": 0.654,
+      "step": 1525
+    },
+    {
+      "epoch": 0.14851581508515815,
+      "grad_norm": 1.2700043251684836,
+      "learning_rate": 9.636548811443685e-06,
+      "loss": 0.3338,
+      "step": 1526
+    },
+    {
+      "epoch": 0.14861313868613138,
+      "grad_norm": 1.4124836827913858,
+      "learning_rate": 9.63595863435359e-06,
+      "loss": 0.3551,
+      "step": 1527
+    },
+    {
+      "epoch": 0.14871046228710463,
+      "grad_norm": 1.3732601776051463,
+      "learning_rate": 9.635367996585436e-06,
+      "loss": 0.4212,
+      "step": 1528
+    },
+    {
+      "epoch": 0.14880778588807786,
+      "grad_norm": 1.4785898006079692,
+      "learning_rate": 9.634776898197916e-06,
+      "loss": 0.416,
+      "step": 1529
+    },
+    {
+      "epoch": 0.14890510948905109,
+      "grad_norm": 1.5889313350171215,
+      "learning_rate": 9.634185339249766e-06,
+      "loss": 0.5277,
+      "step": 1530
+    },
+    {
+      "epoch": 0.14900243309002434,
+      "grad_norm": 1.7475817866143981,
+      "learning_rate": 9.63359331979977e-06,
+      "loss": 0.5202,
+      "step": 1531
+    },
+    {
+      "epoch": 0.14909975669099756,
+      "grad_norm": 1.5329427899001755,
+      "learning_rate": 9.633000839906758e-06,
+      "loss": 0.4283,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1491970802919708,
+      "grad_norm": 1.3789605408265815,
+      "learning_rate": 9.632407899629606e-06,
+      "loss": 0.41,
+      "step": 1533
+    },
+    {
+      "epoch": 0.14929440389294404,
+      "grad_norm": 1.725959361785896,
+      "learning_rate": 9.631814499027233e-06,
+      "loss": 0.6289,
+      "step": 1534
+    },
+    {
+      "epoch": 0.14939172749391727,
+      "grad_norm": 1.5432692609357797,
+      "learning_rate": 9.631220638158605e-06,
+      "loss": 0.5,
+      "step": 1535
+    },
+    {
+      "epoch": 0.14948905109489052,
+      "grad_norm": 1.6556108789068573,
+      "learning_rate": 9.630626317082737e-06,
+      "loss": 0.3819,
+      "step": 1536
+    },
+    {
+      "epoch": 0.14958637469586375,
+      "grad_norm": 1.4498977098887442,
+      "learning_rate": 9.630031535858686e-06,
+      "loss": 0.4317,
+      "step": 1537
+    },
+    {
+      "epoch": 0.14968369829683698,
+      "grad_norm": 1.1232180788321369,
+      "learning_rate": 9.629436294545555e-06,
+      "loss": 0.4004,
+      "step": 1538
+    },
+    {
+      "epoch": 0.14978102189781023,
+      "grad_norm": 0.9949950497949807,
+      "learning_rate": 9.628840593202494e-06,
+      "loss": 0.2008,
+      "step": 1539
+    },
+    {
+      "epoch": 0.14987834549878346,
+      "grad_norm": 1.431426278132333,
+      "learning_rate": 9.628244431888699e-06,
+      "loss": 0.3689,
+      "step": 1540
+    },
+    {
+      "epoch": 0.14997566909975668,
+      "grad_norm": 1.575987397356523,
+      "learning_rate": 9.627647810663407e-06,
+      "loss": 0.5513,
+      "step": 1541
+    },
+    {
+      "epoch": 0.15007299270072993,
+      "grad_norm": 1.5419042077794642,
+      "learning_rate": 9.627050729585911e-06,
+      "loss": 0.4614,
+      "step": 1542
+    },
+    {
+      "epoch": 0.15017031630170316,
+      "grad_norm": 1.6695059275012083,
+      "learning_rate": 9.626453188715539e-06,
+      "loss": 0.5111,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1502676399026764,
+      "grad_norm": 1.5402255238707527,
+      "learning_rate": 9.625855188111668e-06,
+      "loss": 0.4209,
+      "step": 1544
+    },
+    {
+      "epoch": 0.15036496350364964,
+      "grad_norm": 1.4645797288107798,
+      "learning_rate": 9.625256727833726e-06,
+      "loss": 0.4852,
+      "step": 1545
+    },
+    {
+      "epoch": 0.15046228710462287,
+      "grad_norm": 2.0138530187225845,
+      "learning_rate": 9.62465780794118e-06,
+      "loss": 0.4272,
+      "step": 1546
+    },
+    {
+      "epoch": 0.15055961070559612,
+      "grad_norm": 1.7939871096323345,
+      "learning_rate": 9.624058428493543e-06,
+      "loss": 0.3864,
+      "step": 1547
+    },
+    {
+      "epoch": 0.15065693430656935,
+      "grad_norm": 1.5936734798237622,
+      "learning_rate": 9.62345858955038e-06,
+      "loss": 0.5951,
+      "step": 1548
+    },
+    {
+      "epoch": 0.15075425790754257,
+      "grad_norm": 1.381736638575513,
+      "learning_rate": 9.622858291171295e-06,
+      "loss": 0.5078,
+      "step": 1549
+    },
+    {
+      "epoch": 0.15085158150851583,
+      "grad_norm": 1.2680468052820635,
+      "learning_rate": 9.622257533415939e-06,
+      "loss": 0.3314,
+      "step": 1550
+    },
+    {
+      "epoch": 0.15094890510948905,
+      "grad_norm": 1.5886359348363517,
+      "learning_rate": 9.621656316344011e-06,
+      "loss": 0.5985,
+      "step": 1551
+    },
+    {
+      "epoch": 0.15104622871046228,
+      "grad_norm": 1.631001321245941,
+      "learning_rate": 9.621054640015255e-06,
+      "loss": 0.6297,
+      "step": 1552
+    },
+    {
+      "epoch": 0.15114355231143553,
+      "grad_norm": 1.7004985330783402,
+      "learning_rate": 9.62045250448946e-06,
+      "loss": 0.5153,
+      "step": 1553
+    },
+    {
+      "epoch": 0.15124087591240876,
+      "grad_norm": 1.414050528965644,
+      "learning_rate": 9.619849909826457e-06,
+      "loss": 0.2651,
+      "step": 1554
+    },
+    {
+      "epoch": 0.15133819951338198,
+      "grad_norm": 1.3361950007111751,
+      "learning_rate": 9.61924685608613e-06,
+      "loss": 0.4179,
+      "step": 1555
+    },
+    {
+      "epoch": 0.15143552311435524,
+      "grad_norm": 1.2305020766816175,
+      "learning_rate": 9.618643343328404e-06,
+      "loss": 0.3342,
+      "step": 1556
+    },
+    {
+      "epoch": 0.15153284671532846,
+      "grad_norm": 1.3364057110807985,
+      "learning_rate": 9.618039371613251e-06,
+      "loss": 0.357,
+      "step": 1557
+    },
+    {
+      "epoch": 0.15163017031630172,
+      "grad_norm": 0.9846564904659728,
+      "learning_rate": 9.617434941000685e-06,
+      "loss": 0.2278,
+      "step": 1558
+    },
+    {
+      "epoch": 0.15172749391727494,
+      "grad_norm": 1.4874184978820846,
+      "learning_rate": 9.616830051550772e-06,
+      "loss": 0.4467,
+      "step": 1559
+    },
+    {
+      "epoch": 0.15182481751824817,
+      "grad_norm": 1.79907754997464,
+      "learning_rate": 9.61622470332362e-06,
+      "loss": 0.5501,
+      "step": 1560
+    },
+    {
+      "epoch": 0.15192214111922142,
+      "grad_norm": 1.2290536645357835,
+      "learning_rate": 9.61561889637938e-06,
+      "loss": 0.3149,
+      "step": 1561
+    },
+    {
+      "epoch": 0.15201946472019465,
+      "grad_norm": 1.5048179340178087,
+      "learning_rate": 9.615012630778254e-06,
+      "loss": 0.5367,
+      "step": 1562
+    },
+    {
+      "epoch": 0.15211678832116787,
+      "grad_norm": 1.387431259858161,
+      "learning_rate": 9.614405906580486e-06,
+      "loss": 0.4953,
+      "step": 1563
+    },
+    {
+      "epoch": 0.15221411192214113,
+      "grad_norm": 1.4159610711473967,
+      "learning_rate": 9.613798723846368e-06,
+      "loss": 0.454,
+      "step": 1564
+    },
+    {
+      "epoch": 0.15231143552311435,
+      "grad_norm": 1.2005509919566202,
+      "learning_rate": 9.613191082636235e-06,
+      "loss": 0.3945,
+      "step": 1565
+    },
+    {
+      "epoch": 0.15240875912408758,
+      "grad_norm": 1.518451218591156,
+      "learning_rate": 9.612582983010468e-06,
+      "loss": 0.42,
+      "step": 1566
+    },
+    {
+      "epoch": 0.15250608272506083,
+      "grad_norm": 1.2817177267697137,
+      "learning_rate": 9.611974425029494e-06,
+      "loss": 0.4119,
+      "step": 1567
+    },
+    {
+      "epoch": 0.15260340632603406,
+      "grad_norm": 1.3182769071429,
+      "learning_rate": 9.611365408753787e-06,
+      "loss": 0.4301,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1527007299270073,
+      "grad_norm": 1.2668371165350867,
+      "learning_rate": 9.610755934243864e-06,
+      "loss": 0.3415,
+      "step": 1569
+    },
+    {
+      "epoch": 0.15279805352798054,
+      "grad_norm": 1.334265705787435,
+      "learning_rate": 9.610146001560293e-06,
+      "loss": 0.325,
+      "step": 1570
+    },
+    {
+      "epoch": 0.15289537712895376,
+      "grad_norm": 1.405628575667756,
+      "learning_rate": 9.609535610763678e-06,
+      "loss": 0.4,
+      "step": 1571
+    },
+    {
+      "epoch": 0.15299270072992702,
+      "grad_norm": 1.5931859233666277,
+      "learning_rate": 9.608924761914677e-06,
+      "loss": 0.643,
+      "step": 1572
+    },
+    {
+      "epoch": 0.15309002433090024,
+      "grad_norm": 1.323715339329346,
+      "learning_rate": 9.608313455073989e-06,
+      "loss": 0.4832,
+      "step": 1573
+    },
+    {
+      "epoch": 0.15318734793187347,
+      "grad_norm": 1.1603088792271297,
+      "learning_rate": 9.60770169030236e-06,
+      "loss": 0.2684,
+      "step": 1574
+    },
+    {
+      "epoch": 0.15328467153284672,
+      "grad_norm": 1.4578030666688024,
+      "learning_rate": 9.607089467660581e-06,
+      "loss": 0.4418,
+      "step": 1575
+    },
+    {
+      "epoch": 0.15338199513381995,
+      "grad_norm": 1.2739086566679132,
+      "learning_rate": 9.606476787209493e-06,
+      "loss": 0.3847,
+      "step": 1576
+    },
+    {
+      "epoch": 0.15347931873479317,
+      "grad_norm": 1.4031538044918876,
+      "learning_rate": 9.605863649009973e-06,
+      "loss": 0.3672,
+      "step": 1577
+    },
+    {
+      "epoch": 0.15357664233576643,
+      "grad_norm": 1.473592849907526,
+      "learning_rate": 9.605250053122951e-06,
+      "loss": 0.3955,
+      "step": 1578
+    },
+    {
+      "epoch": 0.15367396593673965,
+      "grad_norm": 1.6950520258208177,
+      "learning_rate": 9.604635999609402e-06,
+      "loss": 0.6923,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1537712895377129,
+      "grad_norm": 1.6074239515288835,
+      "learning_rate": 9.604021488530342e-06,
+      "loss": 0.4771,
+      "step": 1580
+    },
+    {
+      "epoch": 0.15386861313868613,
+      "grad_norm": 1.5289432511411145,
+      "learning_rate": 9.603406519946838e-06,
+      "loss": 0.5881,
+      "step": 1581
+    },
+    {
+      "epoch": 0.15396593673965936,
+      "grad_norm": 1.3225323677068181,
+      "learning_rate": 9.602791093919998e-06,
+      "loss": 0.3128,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1540632603406326,
+      "grad_norm": 1.467417498061456,
+      "learning_rate": 9.60217521051098e-06,
+      "loss": 0.545,
+      "step": 1583
+    },
+    {
+      "epoch": 0.15416058394160584,
+      "grad_norm": 1.7568491012309082,
+      "learning_rate": 9.60155886978098e-06,
+      "loss": 0.7054,
+      "step": 1584
+    },
+    {
+      "epoch": 0.15425790754257906,
+      "grad_norm": 1.5606257069028109,
+      "learning_rate": 9.600942071791248e-06,
+      "loss": 0.4329,
+      "step": 1585
+    },
+    {
+      "epoch": 0.15435523114355232,
+      "grad_norm": 1.5727160833264413,
+      "learning_rate": 9.600324816603074e-06,
+      "loss": 0.6128,
+      "step": 1586
+    },
+    {
+      "epoch": 0.15445255474452554,
+      "grad_norm": 1.3864503412663605,
+      "learning_rate": 9.599707104277796e-06,
+      "loss": 0.573,
+      "step": 1587
+    },
+    {
+      "epoch": 0.15454987834549877,
+      "grad_norm": 1.4232761061254342,
+      "learning_rate": 9.599088934876794e-06,
+      "loss": 0.4136,
+      "step": 1588
+    },
+    {
+      "epoch": 0.15464720194647202,
+      "grad_norm": 1.3399427727677786,
+      "learning_rate": 9.598470308461499e-06,
+      "loss": 0.257,
+      "step": 1589
+    },
+    {
+      "epoch": 0.15474452554744525,
+      "grad_norm": 1.61635763649276,
+      "learning_rate": 9.597851225093382e-06,
+      "loss": 0.566,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1548418491484185,
+      "grad_norm": 1.6304164262097627,
+      "learning_rate": 9.597231684833964e-06,
+      "loss": 0.3673,
+      "step": 1591
+    },
+    {
+      "epoch": 0.15493917274939173,
+      "grad_norm": 1.4592987498064005,
+      "learning_rate": 9.596611687744807e-06,
+      "loss": 0.5193,
+      "step": 1592
+    },
+    {
+      "epoch": 0.15503649635036496,
+      "grad_norm": 1.4397292060019447,
+      "learning_rate": 9.595991233887523e-06,
+      "loss": 0.3236,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1551338199513382,
+      "grad_norm": 1.2246835494507005,
+      "learning_rate": 9.595370323323763e-06,
+      "loss": 0.2397,
+      "step": 1594
+    },
+    {
+      "epoch": 0.15523114355231143,
+      "grad_norm": 1.530797619071646,
+      "learning_rate": 9.59474895611523e-06,
+      "loss": 0.4537,
+      "step": 1595
+    },
+    {
+      "epoch": 0.15532846715328466,
+      "grad_norm": 0.9400393110536889,
+      "learning_rate": 9.594127132323669e-06,
+      "loss": 0.1899,
+      "step": 1596
+    },
+    {
+      "epoch": 0.15542579075425791,
+      "grad_norm": 1.167634539806263,
+      "learning_rate": 9.593504852010872e-06,
+      "loss": 0.353,
+      "step": 1597
+    },
+    {
+      "epoch": 0.15552311435523114,
+      "grad_norm": 1.6772160290018319,
+      "learning_rate": 9.592882115238675e-06,
+      "loss": 0.4194,
+      "step": 1598
+    },
+    {
+      "epoch": 0.15562043795620437,
+      "grad_norm": 1.4391641520861267,
+      "learning_rate": 9.592258922068958e-06,
+      "loss": 0.4767,
+      "step": 1599
+    },
+    {
+      "epoch": 0.15571776155717762,
+      "grad_norm": 1.544673007447179,
+      "learning_rate": 9.591635272563648e-06,
+      "loss": 0.3175,
+      "step": 1600
+    },
+    {
+      "epoch": 0.15581508515815085,
+      "grad_norm": 1.4189512773822923,
+      "learning_rate": 9.591011166784721e-06,
+      "loss": 0.4834,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1559124087591241,
+      "grad_norm": 1.2414753149853184,
+      "learning_rate": 9.590386604794191e-06,
+      "loss": 0.3657,
+      "step": 1602
+    },
+    {
+      "epoch": 0.15600973236009733,
+      "grad_norm": 1.0236785255419305,
+      "learning_rate": 9.589761586654122e-06,
+      "loss": 0.2011,
+      "step": 1603
+    },
+    {
+      "epoch": 0.15610705596107055,
+      "grad_norm": 2.3461369884265357,
+      "learning_rate": 9.589136112426625e-06,
+      "loss": 0.4024,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1562043795620438,
+      "grad_norm": 1.2849479900774115,
+      "learning_rate": 9.588510182173851e-06,
+      "loss": 0.3527,
+      "step": 1605
+    },
+    {
+      "epoch": 0.15630170316301703,
+      "grad_norm": 1.4153286655317308,
+      "learning_rate": 9.587883795958001e-06,
+      "loss": 0.4149,
+      "step": 1606
+    },
+    {
+      "epoch": 0.15639902676399026,
+      "grad_norm": 1.6599870662874754,
+      "learning_rate": 9.587256953841317e-06,
+      "loss": 0.6479,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1564963503649635,
+      "grad_norm": 1.6670860080877101,
+      "learning_rate": 9.58662965588609e-06,
+      "loss": 0.5825,
+      "step": 1608
+    },
+    {
+      "epoch": 0.15659367396593674,
+      "grad_norm": 1.7776280437765584,
+      "learning_rate": 9.586001902154655e-06,
+      "loss": 0.5798,
+      "step": 1609
+    },
+    {
+      "epoch": 0.15669099756690996,
+      "grad_norm": 1.5456297515043347,
+      "learning_rate": 9.585373692709391e-06,
+      "loss": 0.4583,
+      "step": 1610
+    },
+    {
+      "epoch": 0.15678832116788322,
+      "grad_norm": 1.8806083091738082,
+      "learning_rate": 9.584745027612728e-06,
+      "loss": 0.4736,
+      "step": 1611
+    },
+    {
+      "epoch": 0.15688564476885644,
+      "grad_norm": 1.4790926453601037,
+      "learning_rate": 9.584115906927131e-06,
+      "loss": 0.4172,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1569829683698297,
+      "grad_norm": 3.3021500316987633,
+      "learning_rate": 9.58348633071512e-06,
+      "loss": 0.472,
+      "step": 1613
+    },
+    {
+      "epoch": 0.15708029197080292,
+      "grad_norm": 1.860435632122749,
+      "learning_rate": 9.582856299039253e-06,
+      "loss": 0.4743,
+      "step": 1614
+    },
+    {
+      "epoch": 0.15717761557177615,
+      "grad_norm": 1.7557070181222967,
+      "learning_rate": 9.58222581196214e-06,
+      "loss": 0.2907,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1572749391727494,
+      "grad_norm": 1.5588238003780286,
+      "learning_rate": 9.581594869546433e-06,
+      "loss": 0.3803,
+      "step": 1616
+    },
+    {
+      "epoch": 0.15737226277372263,
+      "grad_norm": 1.5265824940366777,
+      "learning_rate": 9.580963471854825e-06,
+      "loss": 0.3163,
+      "step": 1617
+    },
+    {
+      "epoch": 0.15746958637469585,
+      "grad_norm": 1.5425233608560427,
+      "learning_rate": 9.580331618950063e-06,
+      "loss": 0.3884,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1575669099756691,
+      "grad_norm": 1.4123635386488018,
+      "learning_rate": 9.579699310894932e-06,
+      "loss": 0.382,
+      "step": 1619
+    },
+    {
+      "epoch": 0.15766423357664233,
+      "grad_norm": 1.578019469103596,
+      "learning_rate": 9.579066547752266e-06,
+      "loss": 0.4293,
+      "step": 1620
+    },
+    {
+      "epoch": 0.15776155717761559,
+      "grad_norm": 1.6566990657429592,
+      "learning_rate": 9.578433329584943e-06,
+      "loss": 0.2878,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1578588807785888,
+      "grad_norm": 1.5290043771605026,
+      "learning_rate": 9.577799656455886e-06,
+      "loss": 0.4483,
+      "step": 1622
+    },
+    {
+      "epoch": 0.15795620437956204,
+      "grad_norm": 1.7268752423292135,
+      "learning_rate": 9.577165528428063e-06,
+      "loss": 0.4805,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1580535279805353,
+      "grad_norm": 1.3495189675110832,
+      "learning_rate": 9.576530945564488e-06,
+      "loss": 0.3161,
+      "step": 1624
+    },
+    {
+      "epoch": 0.15815085158150852,
+      "grad_norm": 1.4763829359235794,
+      "learning_rate": 9.575895907928218e-06,
+      "loss": 0.4825,
+      "step": 1625
+    },
+    {
+      "epoch": 0.15824817518248174,
+      "grad_norm": 1.686991367686583,
+      "learning_rate": 9.575260415582362e-06,
+      "loss": 0.3016,
+      "step": 1626
+    },
+    {
+      "epoch": 0.158345498783455,
+      "grad_norm": 1.3390220591470878,
+      "learning_rate": 9.574624468590065e-06,
+      "loss": 0.4523,
+      "step": 1627
+    },
+    {
+      "epoch": 0.15844282238442822,
+      "grad_norm": 1.8698808087393168,
+      "learning_rate": 9.573988067014523e-06,
+      "loss": 0.5203,
+      "step": 1628
+    },
+    {
+      "epoch": 0.15854014598540145,
+      "grad_norm": 1.4032165021732874,
+      "learning_rate": 9.573351210918976e-06,
+      "loss": 0.3678,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1586374695863747,
+      "grad_norm": 1.4017015011859046,
+      "learning_rate": 9.572713900366707e-06,
+      "loss": 0.2798,
+      "step": 1630
+    },
+    {
+      "epoch": 0.15873479318734793,
+      "grad_norm": 1.4441030854971395,
+      "learning_rate": 9.572076135421048e-06,
+      "loss": 0.3514,
+      "step": 1631
+    },
+    {
+      "epoch": 0.15883211678832118,
+      "grad_norm": 1.3629792761623065,
+      "learning_rate": 9.571437916145373e-06,
+      "loss": 0.4604,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1589294403892944,
+      "grad_norm": 1.376972344446985,
+      "learning_rate": 9.570799242603101e-06,
+      "loss": 0.4603,
+      "step": 1633
+    },
+    {
+      "epoch": 0.15902676399026763,
+      "grad_norm": 1.5637421057827365,
+      "learning_rate": 9.5701601148577e-06,
+      "loss": 0.5575,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1591240875912409,
+      "grad_norm": 1.4338457681188446,
+      "learning_rate": 9.56952053297268e-06,
+      "loss": 0.532,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1592214111922141,
+      "grad_norm": 1.4858651962900338,
+      "learning_rate": 9.568880497011597e-06,
+      "loss": 0.4951,
+      "step": 1636
+    },
+    {
+      "epoch": 0.15931873479318734,
+      "grad_norm": 1.543423201839799,
+      "learning_rate": 9.568240007038048e-06,
+      "loss": 0.5278,
+      "step": 1637
+    },
+    {
+      "epoch": 0.1594160583941606,
+      "grad_norm": 1.408319688012345,
+      "learning_rate": 9.567599063115683e-06,
+      "loss": 0.4474,
+      "step": 1638
+    },
+    {
+      "epoch": 0.15951338199513382,
+      "grad_norm": 1.2680346779127702,
+      "learning_rate": 9.566957665308192e-06,
+      "loss": 0.3351,
+      "step": 1639
+    },
+    {
+      "epoch": 0.15961070559610704,
+      "grad_norm": 1.6277797838197976,
+      "learning_rate": 9.56631581367931e-06,
+      "loss": 0.3966,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1597080291970803,
+      "grad_norm": 1.5248977314161354,
+      "learning_rate": 9.565673508292818e-06,
+      "loss": 0.5211,
+      "step": 1641
+    },
+    {
+      "epoch": 0.15980535279805352,
+      "grad_norm": 1.7164012466100764,
+      "learning_rate": 9.565030749212546e-06,
+      "loss": 0.5428,
+      "step": 1642
+    },
+    {
+      "epoch": 0.15990267639902678,
+      "grad_norm": 1.6687081549609284,
+      "learning_rate": 9.56438753650236e-06,
+      "loss": 0.2936,
+      "step": 1643
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5678110268585723,
+      "learning_rate": 9.56374387022618e-06,
+      "loss": 0.5166,
+      "step": 1644
+    },
+    {
+      "epoch": 0.16009732360097323,
+      "grad_norm": 1.6983019931785335,
+      "learning_rate": 9.563099750447966e-06,
+      "loss": 0.4822,
+      "step": 1645
+    },
+    {
+      "epoch": 0.16019464720194648,
+      "grad_norm": 1.4431824530543444,
+      "learning_rate": 9.562455177231726e-06,
+      "loss": 0.3212,
+      "step": 1646
+    },
+    {
+      "epoch": 0.1602919708029197,
+      "grad_norm": 3.712828208723791,
+      "learning_rate": 9.56181015064151e-06,
+      "loss": 0.4286,
+      "step": 1647
+    },
+    {
+      "epoch": 0.16038929440389293,
+      "grad_norm": 1.4388083433357408,
+      "learning_rate": 9.561164670741416e-06,
+      "loss": 0.3757,
+      "step": 1648
+    },
+    {
+      "epoch": 0.1604866180048662,
+      "grad_norm": 2.4878081586110117,
+      "learning_rate": 9.560518737595586e-06,
+      "loss": 0.3494,
+      "step": 1649
+    },
+    {
+      "epoch": 0.16058394160583941,
+      "grad_norm": 2.3091262745384706,
+      "learning_rate": 9.559872351268205e-06,
+      "loss": 0.4607,
+      "step": 1650
+    },
+    {
+      "epoch": 0.16068126520681264,
+      "grad_norm": 1.6632563827899045,
+      "learning_rate": 9.559225511823504e-06,
+      "loss": 0.5718,
+      "step": 1651
+    },
+    {
+      "epoch": 0.1607785888077859,
+      "grad_norm": 1.6138862417611177,
+      "learning_rate": 9.558578219325763e-06,
+      "loss": 0.325,
+      "step": 1652
+    },
+    {
+      "epoch": 0.16087591240875912,
+      "grad_norm": 1.1933317040764397,
+      "learning_rate": 9.557930473839303e-06,
+      "loss": 0.339,
+      "step": 1653
+    },
+    {
+      "epoch": 0.16097323600973237,
+      "grad_norm": 0.9728312200944081,
+      "learning_rate": 9.55728227542849e-06,
+      "loss": 0.2395,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1610705596107056,
+      "grad_norm": 1.5521742092214053,
+      "learning_rate": 9.556633624157735e-06,
+      "loss": 0.4613,
+      "step": 1655
+    },
+    {
+      "epoch": 0.16116788321167883,
+      "grad_norm": 1.639740187603822,
+      "learning_rate": 9.555984520091497e-06,
+      "loss": 0.5146,
+      "step": 1656
+    },
+    {
+      "epoch": 0.16126520681265208,
+      "grad_norm": 1.5387772039120604,
+      "learning_rate": 9.555334963294277e-06,
+      "loss": 0.4879,
+      "step": 1657
+    },
+    {
+      "epoch": 0.1613625304136253,
+      "grad_norm": 1.2788374913210725,
+      "learning_rate": 9.554684953830622e-06,
+      "loss": 0.2115,
+      "step": 1658
+    },
+    {
+      "epoch": 0.16145985401459853,
+      "grad_norm": 1.2466060338770748,
+      "learning_rate": 9.554034491765123e-06,
+      "loss": 0.4057,
+      "step": 1659
+    },
+    {
+      "epoch": 0.16155717761557178,
+      "grad_norm": 1.3626765355526065,
+      "learning_rate": 9.553383577162418e-06,
+      "loss": 0.3922,
+      "step": 1660
+    },
+    {
+      "epoch": 0.161654501216545,
+      "grad_norm": 1.4993759287568524,
+      "learning_rate": 9.552732210087188e-06,
+      "loss": 0.5101,
+      "step": 1661
+    },
+    {
+      "epoch": 0.16175182481751824,
+      "grad_norm": 1.4132678080310175,
+      "learning_rate": 9.55208039060416e-06,
+      "loss": 0.4098,
+      "step": 1662
+    },
+    {
+      "epoch": 0.1618491484184915,
+      "grad_norm": 1.3072203759845393,
+      "learning_rate": 9.551428118778105e-06,
+      "loss": 0.4437,
+      "step": 1663
+    },
+    {
+      "epoch": 0.16194647201946472,
+      "grad_norm": 1.4197615961970556,
+      "learning_rate": 9.550775394673841e-06,
+      "loss": 0.4855,
+      "step": 1664
+    },
+    {
+      "epoch": 0.16204379562043797,
+      "grad_norm": 1.1443578178578404,
+      "learning_rate": 9.550122218356228e-06,
+      "loss": 0.2651,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1621411192214112,
+      "grad_norm": 1.6274953169982382,
+      "learning_rate": 9.549468589890173e-06,
+      "loss": 0.5702,
+      "step": 1666
+    },
+    {
+      "epoch": 0.16223844282238442,
+      "grad_norm": 1.5542252970145625,
+      "learning_rate": 9.548814509340631e-06,
+      "loss": 0.3618,
+      "step": 1667
+    },
+    {
+      "epoch": 0.16233576642335767,
+      "grad_norm": 1.5872588267319008,
+      "learning_rate": 9.548159976772593e-06,
+      "loss": 0.5261,
+      "step": 1668
+    },
+    {
+      "epoch": 0.1624330900243309,
+      "grad_norm": 1.1735078752446053,
+      "learning_rate": 9.547504992251102e-06,
+      "loss": 0.2709,
+      "step": 1669
+    },
+    {
+      "epoch": 0.16253041362530413,
+      "grad_norm": 1.8057871189139236,
+      "learning_rate": 9.546849555841247e-06,
+      "loss": 0.3383,
+      "step": 1670
+    },
+    {
+      "epoch": 0.16262773722627738,
+      "grad_norm": 1.4181568031561294,
+      "learning_rate": 9.546193667608155e-06,
+      "loss": 0.4654,
+      "step": 1671
+    },
+    {
+      "epoch": 0.1627250608272506,
+      "grad_norm": 1.3372190697374011,
+      "learning_rate": 9.545537327617004e-06,
+      "loss": 0.4098,
+      "step": 1672
+    },
+    {
+      "epoch": 0.16282238442822383,
+      "grad_norm": 1.4054977948345526,
+      "learning_rate": 9.544880535933015e-06,
+      "loss": 0.488,
+      "step": 1673
+    },
+    {
+      "epoch": 0.16291970802919709,
+      "grad_norm": 1.8103202340533562,
+      "learning_rate": 9.544223292621456e-06,
+      "loss": 0.2989,
+      "step": 1674
+    },
+    {
+      "epoch": 0.1630170316301703,
+      "grad_norm": 1.4424657055300307,
+      "learning_rate": 9.543565597747633e-06,
+      "loss": 0.3545,
+      "step": 1675
+    },
+    {
+      "epoch": 0.16311435523114357,
+      "grad_norm": 1.712897793310079,
+      "learning_rate": 9.542907451376904e-06,
+      "loss": 0.4372,
+      "step": 1676
+    },
+    {
+      "epoch": 0.1632116788321168,
+      "grad_norm": 1.5856342495538354,
+      "learning_rate": 9.542248853574669e-06,
+      "loss": 0.3552,
+      "step": 1677
+    },
+    {
+      "epoch": 0.16330900243309002,
+      "grad_norm": 1.6070757988154845,
+      "learning_rate": 9.541589804406373e-06,
+      "loss": 0.6297,
+      "step": 1678
+    },
+    {
+      "epoch": 0.16340632603406327,
+      "grad_norm": 1.4030835423791206,
+      "learning_rate": 9.540930303937508e-06,
+      "loss": 0.5304,
+      "step": 1679
+    },
+    {
+      "epoch": 0.1635036496350365,
+      "grad_norm": 1.1629420270697914,
+      "learning_rate": 9.540270352233607e-06,
+      "loss": 0.3196,
+      "step": 1680
+    },
+    {
+      "epoch": 0.16360097323600972,
+      "grad_norm": 1.6438421767465334,
+      "learning_rate": 9.53960994936025e-06,
+      "loss": 0.5718,
+      "step": 1681
+    },
+    {
+      "epoch": 0.16369829683698298,
+      "grad_norm": 1.4972655485667212,
+      "learning_rate": 9.538949095383064e-06,
+      "loss": 0.5411,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1637956204379562,
+      "grad_norm": 1.6855463092047138,
+      "learning_rate": 9.538287790367715e-06,
+      "loss": 0.4072,
+      "step": 1683
+    },
+    {
+      "epoch": 0.16389294403892943,
+      "grad_norm": 1.3024464622228382,
+      "learning_rate": 9.537626034379918e-06,
+      "loss": 0.3779,
+      "step": 1684
+    },
+    {
+      "epoch": 0.16399026763990268,
+      "grad_norm": 1.295189693137423,
+      "learning_rate": 9.536963827485435e-06,
+      "loss": 0.3687,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1640875912408759,
+      "grad_norm": 1.4535138830119652,
+      "learning_rate": 9.536301169750068e-06,
+      "loss": 0.4548,
+      "step": 1686
+    },
+    {
+      "epoch": 0.16418491484184916,
+      "grad_norm": 1.199213729997,
+      "learning_rate": 9.535638061239663e-06,
+      "loss": 0.2053,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1642822384428224,
+      "grad_norm": 1.5567691993981325,
+      "learning_rate": 9.534974502020117e-06,
+      "loss": 0.4098,
+      "step": 1688
+    },
+    {
+      "epoch": 0.1643795620437956,
+      "grad_norm": 1.5701473016338705,
+      "learning_rate": 9.534310492157368e-06,
+      "loss": 0.4663,
+      "step": 1689
+    },
+    {
+      "epoch": 0.16447688564476887,
+      "grad_norm": 1.4652608455665965,
+      "learning_rate": 9.533646031717398e-06,
+      "loss": 0.423,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1645742092457421,
+      "grad_norm": 1.556818972222242,
+      "learning_rate": 9.532981120766235e-06,
+      "loss": 0.5823,
+      "step": 1691
+    },
+    {
+      "epoch": 0.16467153284671532,
+      "grad_norm": 1.3176167070500389,
+      "learning_rate": 9.532315759369953e-06,
+      "loss": 0.3369,
+      "step": 1692
+    },
+    {
+      "epoch": 0.16476885644768857,
+      "grad_norm": 1.710131590392248,
+      "learning_rate": 9.531649947594668e-06,
+      "loss": 0.6235,
+      "step": 1693
+    },
+    {
+      "epoch": 0.1648661800486618,
+      "grad_norm": 1.316452070848038,
+      "learning_rate": 9.53098368550654e-06,
+      "loss": 0.2773,
+      "step": 1694
+    },
+    {
+      "epoch": 0.16496350364963502,
+      "grad_norm": 1.3144552108952152,
+      "learning_rate": 9.53031697317178e-06,
+      "loss": 0.4008,
+      "step": 1695
+    },
+    {
+      "epoch": 0.16506082725060828,
+      "grad_norm": 1.6242845867808264,
+      "learning_rate": 9.529649810656638e-06,
+      "loss": 0.4994,
+      "step": 1696
+    },
+    {
+      "epoch": 0.1651581508515815,
+      "grad_norm": 1.285181340955318,
+      "learning_rate": 9.52898219802741e-06,
+      "loss": 0.3565,
+      "step": 1697
+    },
+    {
+      "epoch": 0.16525547445255476,
+      "grad_norm": 1.5859120183692204,
+      "learning_rate": 9.528314135350439e-06,
+      "loss": 0.6057,
+      "step": 1698
+    },
+    {
+      "epoch": 0.16535279805352798,
+      "grad_norm": 1.2413369391689792,
+      "learning_rate": 9.527645622692105e-06,
+      "loss": 0.2912,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1654501216545012,
+      "grad_norm": 1.5626898078072964,
+      "learning_rate": 9.526976660118846e-06,
+      "loss": 0.4912,
+      "step": 1700
+    },
+    {
+      "epoch": 0.16554744525547446,
+      "grad_norm": 1.355302168314411,
+      "learning_rate": 9.526307247697133e-06,
+      "loss": 0.4066,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1656447688564477,
+      "grad_norm": 1.6754743388370108,
+      "learning_rate": 9.525637385493485e-06,
+      "loss": 0.4402,
+      "step": 1702
+    },
+    {
+      "epoch": 0.16574209245742091,
+      "grad_norm": 1.4378330010865907,
+      "learning_rate": 9.524967073574468e-06,
+      "loss": 0.3896,
+      "step": 1703
+    },
+    {
+      "epoch": 0.16583941605839417,
+      "grad_norm": 1.5562357645264613,
+      "learning_rate": 9.524296312006696e-06,
+      "loss": 0.7178,
+      "step": 1704
+    },
+    {
+      "epoch": 0.1659367396593674,
+      "grad_norm": 1.4997676033555023,
+      "learning_rate": 9.523625100856814e-06,
+      "loss": 0.5203,
+      "step": 1705
+    },
+    {
+      "epoch": 0.16603406326034062,
+      "grad_norm": 1.39039181243628,
+      "learning_rate": 9.522953440191528e-06,
+      "loss": 0.4804,
+      "step": 1706
+    },
+    {
+      "epoch": 0.16613138686131387,
+      "grad_norm": 1.2594698773182105,
+      "learning_rate": 9.522281330077579e-06,
+      "loss": 0.31,
+      "step": 1707
+    },
+    {
+      "epoch": 0.1662287104622871,
+      "grad_norm": 1.5394103920539104,
+      "learning_rate": 9.521608770581751e-06,
+      "loss": 0.4579,
+      "step": 1708
+    },
+    {
+      "epoch": 0.16632603406326035,
+      "grad_norm": 1.4703967014570463,
+      "learning_rate": 9.520935761770885e-06,
+      "loss": 0.4732,
+      "step": 1709
+    },
+    {
+      "epoch": 0.16642335766423358,
+      "grad_norm": 1.0444153315520046,
+      "learning_rate": 9.520262303711851e-06,
+      "loss": 0.2468,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1665206812652068,
+      "grad_norm": 1.4440019594110525,
+      "learning_rate": 9.519588396471572e-06,
+      "loss": 0.4979,
+      "step": 1711
+    },
+    {
+      "epoch": 0.16661800486618006,
+      "grad_norm": 1.6467368949298022,
+      "learning_rate": 9.518914040117018e-06,
+      "loss": 0.603,
+      "step": 1712
+    },
+    {
+      "epoch": 0.16671532846715328,
+      "grad_norm": 1.656027868957794,
+      "learning_rate": 9.518239234715198e-06,
+      "loss": 0.3534,
+      "step": 1713
+    },
+    {
+      "epoch": 0.1668126520681265,
+      "grad_norm": 1.409360793352949,
+      "learning_rate": 9.517563980333169e-06,
+      "loss": 0.4442,
+      "step": 1714
+    },
+    {
+      "epoch": 0.16690997566909976,
+      "grad_norm": 1.4429795690770129,
+      "learning_rate": 9.51688827703803e-06,
+      "loss": 0.4347,
+      "step": 1715
+    },
+    {
+      "epoch": 0.167007299270073,
+      "grad_norm": 1.2256612199861667,
+      "learning_rate": 9.516212124896926e-06,
+      "loss": 0.3582,
+      "step": 1716
+    },
+    {
+      "epoch": 0.16710462287104622,
+      "grad_norm": 1.340106815948813,
+      "learning_rate": 9.515535523977047e-06,
+      "loss": 0.4494,
+      "step": 1717
+    },
+    {
+      "epoch": 0.16720194647201947,
+      "grad_norm": 1.8033632646616307,
+      "learning_rate": 9.514858474345628e-06,
+      "loss": 0.7254,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1672992700729927,
+      "grad_norm": 1.461471704742246,
+      "learning_rate": 9.514180976069948e-06,
+      "loss": 0.4431,
+      "step": 1719
+    },
+    {
+      "epoch": 0.16739659367396595,
+      "grad_norm": 1.8149337871023152,
+      "learning_rate": 9.513503029217329e-06,
+      "loss": 0.6808,
+      "step": 1720
+    },
+    {
+      "epoch": 0.16749391727493917,
+      "grad_norm": 1.4317488687976054,
+      "learning_rate": 9.51282463385514e-06,
+      "loss": 0.3969,
+      "step": 1721
+    },
+    {
+      "epoch": 0.1675912408759124,
+      "grad_norm": 1.406660867644435,
+      "learning_rate": 9.512145790050793e-06,
+      "loss": 0.4466,
+      "step": 1722
+    },
+    {
+      "epoch": 0.16768856447688565,
+      "grad_norm": 1.5087949092220858,
+      "learning_rate": 9.511466497871747e-06,
+      "loss": 0.3588,
+      "step": 1723
+    },
+    {
+      "epoch": 0.16778588807785888,
+      "grad_norm": 1.3780878680496882,
+      "learning_rate": 9.5107867573855e-06,
+      "loss": 0.4136,
+      "step": 1724
+    },
+    {
+      "epoch": 0.1678832116788321,
+      "grad_norm": 1.1785521443758606,
+      "learning_rate": 9.510106568659601e-06,
+      "loss": 0.3319,
+      "step": 1725
+    },
+    {
+      "epoch": 0.16798053527980536,
+      "grad_norm": 1.4150065437408217,
+      "learning_rate": 9.50942593176164e-06,
+      "loss": 0.3619,
+      "step": 1726
+    },
+    {
+      "epoch": 0.16807785888077859,
+      "grad_norm": 1.5810685607791577,
+      "learning_rate": 9.508744846759254e-06,
+      "loss": 0.5204,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1681751824817518,
+      "grad_norm": 1.5507123725258296,
+      "learning_rate": 9.50806331372012e-06,
+      "loss": 0.3017,
+      "step": 1728
+    },
+    {
+      "epoch": 0.16827250608272507,
+      "grad_norm": 1.7448176899198176,
+      "learning_rate": 9.507381332711963e-06,
+      "loss": 0.6488,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1683698296836983,
+      "grad_norm": 2.0203041353812243,
+      "learning_rate": 9.506698903802553e-06,
+      "loss": 0.2868,
+      "step": 1730
+    },
+    {
+      "epoch": 0.16846715328467154,
+      "grad_norm": 1.425557408986151,
+      "learning_rate": 9.506016027059703e-06,
+      "loss": 0.4181,
+      "step": 1731
+    },
+    {
+      "epoch": 0.16856447688564477,
+      "grad_norm": 1.658389742609111,
+      "learning_rate": 9.505332702551272e-06,
+      "loss": 0.4834,
+      "step": 1732
+    },
+    {
+      "epoch": 0.168661800486618,
+      "grad_norm": 1.6313220070332846,
+      "learning_rate": 9.50464893034516e-06,
+      "loss": 0.6351,
+      "step": 1733
+    },
+    {
+      "epoch": 0.16875912408759125,
+      "grad_norm": 1.4860828412814417,
+      "learning_rate": 9.503964710509314e-06,
+      "loss": 0.384,
+      "step": 1734
+    },
+    {
+      "epoch": 0.16885644768856448,
+      "grad_norm": 1.5665989326823084,
+      "learning_rate": 9.503280043111729e-06,
+      "loss": 0.5031,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1689537712895377,
+      "grad_norm": 1.2627591310970376,
+      "learning_rate": 9.502594928220437e-06,
+      "loss": 0.3557,
+      "step": 1736
+    },
+    {
+      "epoch": 0.16905109489051096,
+      "grad_norm": 1.6101827723851228,
+      "learning_rate": 9.50190936590352e-06,
+      "loss": 0.3886,
+      "step": 1737
+    },
+    {
+      "epoch": 0.16914841849148418,
+      "grad_norm": 1.190927027644026,
+      "learning_rate": 9.5012233562291e-06,
+      "loss": 0.3,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1692457420924574,
+      "grad_norm": 1.6452233677093766,
+      "learning_rate": 9.50053689926535e-06,
+      "loss": 0.5808,
+      "step": 1739
+    },
+    {
+      "epoch": 0.16934306569343066,
+      "grad_norm": 1.607284224817037,
+      "learning_rate": 9.499849995080482e-06,
+      "loss": 0.5726,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1694403892944039,
+      "grad_norm": 1.360873175063302,
+      "learning_rate": 9.499162643742754e-06,
+      "loss": 0.3294,
+      "step": 1741
+    },
+    {
+      "epoch": 0.16953771289537714,
+      "grad_norm": 1.6205396325650636,
+      "learning_rate": 9.49847484532047e-06,
+      "loss": 0.5496,
+      "step": 1742
+    },
+    {
+      "epoch": 0.16963503649635037,
+      "grad_norm": 1.6677491090337848,
+      "learning_rate": 9.497786599881973e-06,
+      "loss": 0.5745,
+      "step": 1743
+    },
+    {
+      "epoch": 0.1697323600973236,
+      "grad_norm": 1.4765151889225172,
+      "learning_rate": 9.497097907495658e-06,
+      "loss": 0.3552,
+      "step": 1744
+    },
+    {
+      "epoch": 0.16982968369829685,
+      "grad_norm": 1.4991516257283077,
+      "learning_rate": 9.496408768229962e-06,
+      "loss": 0.6004,
+      "step": 1745
+    },
+    {
+      "epoch": 0.16992700729927007,
+      "grad_norm": 1.394241003611109,
+      "learning_rate": 9.49571918215336e-06,
+      "loss": 0.4166,
+      "step": 1746
+    },
+    {
+      "epoch": 0.1700243309002433,
+      "grad_norm": 1.2418310265706307,
+      "learning_rate": 9.495029149334381e-06,
+      "loss": 0.3754,
+      "step": 1747
+    },
+    {
+      "epoch": 0.17012165450121655,
+      "grad_norm": 1.7344174079178016,
+      "learning_rate": 9.494338669841592e-06,
+      "loss": 0.6136,
+      "step": 1748
+    },
+    {
+      "epoch": 0.17021897810218978,
+      "grad_norm": 1.689754745813109,
+      "learning_rate": 9.493647743743605e-06,
+      "loss": 0.3066,
+      "step": 1749
+    },
+    {
+      "epoch": 0.170316301703163,
+      "grad_norm": 1.5986274434851808,
+      "learning_rate": 9.492956371109083e-06,
+      "loss": 0.6476,
+      "step": 1750
+    },
+    {
+      "epoch": 0.17041362530413626,
+      "grad_norm": 1.3892856963539753,
+      "learning_rate": 9.492264552006725e-06,
+      "loss": 0.2438,
+      "step": 1751
+    },
+    {
+      "epoch": 0.17051094890510948,
+      "grad_norm": 1.3744062095245357,
+      "learning_rate": 9.491572286505275e-06,
+      "loss": 0.4154,
+      "step": 1752
+    },
+    {
+      "epoch": 0.17060827250608274,
+      "grad_norm": 1.3041989445373636,
+      "learning_rate": 9.490879574673528e-06,
+      "loss": 0.3603,
+      "step": 1753
+    },
+    {
+      "epoch": 0.17070559610705596,
+      "grad_norm": 1.2198251236981021,
+      "learning_rate": 9.490186416580317e-06,
+      "loss": 0.3382,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1708029197080292,
+      "grad_norm": 1.0699077871285796,
+      "learning_rate": 9.489492812294521e-06,
+      "loss": 0.2805,
+      "step": 1755
+    },
+    {
+      "epoch": 0.17090024330900244,
+      "grad_norm": 1.8289792925797566,
+      "learning_rate": 9.488798761885064e-06,
+      "loss": 0.2551,
+      "step": 1756
+    },
+    {
+      "epoch": 0.17099756690997567,
+      "grad_norm": 1.5156970449411904,
+      "learning_rate": 9.488104265420917e-06,
+      "loss": 0.5468,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1710948905109489,
+      "grad_norm": 1.3669899498040559,
+      "learning_rate": 9.487409322971089e-06,
+      "loss": 0.4705,
+      "step": 1758
+    },
+    {
+      "epoch": 0.17119221411192215,
+      "grad_norm": 1.4212977316967985,
+      "learning_rate": 9.486713934604638e-06,
+      "loss": 0.5259,
+      "step": 1759
+    },
+    {
+      "epoch": 0.17128953771289537,
+      "grad_norm": 1.3256503218660822,
+      "learning_rate": 9.486018100390668e-06,
+      "loss": 0.3825,
+      "step": 1760
+    },
+    {
+      "epoch": 0.17138686131386863,
+      "grad_norm": 1.3448672418414023,
+      "learning_rate": 9.485321820398321e-06,
+      "loss": 0.4984,
+      "step": 1761
+    },
+    {
+      "epoch": 0.17148418491484185,
+      "grad_norm": 1.3293122762885854,
+      "learning_rate": 9.484625094696788e-06,
+      "loss": 0.4419,
+      "step": 1762
+    },
+    {
+      "epoch": 0.17158150851581508,
+      "grad_norm": 1.5749728003681251,
+      "learning_rate": 9.483927923355303e-06,
+      "loss": 0.4512,
+      "step": 1763
+    },
+    {
+      "epoch": 0.17167883211678833,
+      "grad_norm": 2.1875449039755,
+      "learning_rate": 9.483230306443144e-06,
+      "loss": 0.4606,
+      "step": 1764
+    },
+    {
+      "epoch": 0.17177615571776156,
+      "grad_norm": 1.4675466599593059,
+      "learning_rate": 9.482532244029632e-06,
+      "loss": 0.5098,
+      "step": 1765
+    },
+    {
+      "epoch": 0.17187347931873478,
+      "grad_norm": 1.4395657499189969,
+      "learning_rate": 9.481833736184137e-06,
+      "loss": 0.4196,
+      "step": 1766
+    },
+    {
+      "epoch": 0.17197080291970804,
+      "grad_norm": 1.6202346179751734,
+      "learning_rate": 9.48113478297607e-06,
+      "loss": 0.4083,
+      "step": 1767
+    },
+    {
+      "epoch": 0.17206812652068126,
+      "grad_norm": 1.943359375,
+      "learning_rate": 9.480435384474884e-06,
+      "loss": 0.3829,
+      "step": 1768
+    },
+    {
+      "epoch": 0.1721654501216545,
+      "grad_norm": 1.3957800309361543,
+      "learning_rate": 9.47973554075008e-06,
+      "loss": 0.4776,
+      "step": 1769
+    },
+    {
+      "epoch": 0.17226277372262774,
+      "grad_norm": 1.3277740014703983,
+      "learning_rate": 9.479035251871202e-06,
+      "loss": 0.2944,
+      "step": 1770
+    },
+    {
+      "epoch": 0.17236009732360097,
+      "grad_norm": 1.5955109684829234,
+      "learning_rate": 9.478334517907838e-06,
+      "loss": 0.4713,
+      "step": 1771
+    },
+    {
+      "epoch": 0.17245742092457422,
+      "grad_norm": 1.209763677864614,
+      "learning_rate": 9.477633338929621e-06,
+      "loss": 0.2925,
+      "step": 1772
+    },
+    {
+      "epoch": 0.17255474452554745,
+      "grad_norm": 1.6082316661319236,
+      "learning_rate": 9.476931715006225e-06,
+      "loss": 0.6037,
+      "step": 1773
+    },
+    {
+      "epoch": 0.17265206812652067,
+      "grad_norm": 1.5310145605828824,
+      "learning_rate": 9.476229646207375e-06,
+      "loss": 0.427,
+      "step": 1774
+    },
+    {
+      "epoch": 0.17274939172749393,
+      "grad_norm": 1.5161322305327478,
+      "learning_rate": 9.475527132602833e-06,
+      "loss": 0.5765,
+      "step": 1775
+    },
+    {
+      "epoch": 0.17284671532846715,
+      "grad_norm": 1.5515912532543141,
+      "learning_rate": 9.47482417426241e-06,
+      "loss": 0.4693,
+      "step": 1776
+    },
+    {
+      "epoch": 0.17294403892944038,
+      "grad_norm": 1.273583152257964,
+      "learning_rate": 9.474120771255956e-06,
+      "loss": 0.401,
+      "step": 1777
+    },
+    {
+      "epoch": 0.17304136253041363,
+      "grad_norm": 1.3058387108850102,
+      "learning_rate": 9.473416923653373e-06,
+      "loss": 0.4651,
+      "step": 1778
+    },
+    {
+      "epoch": 0.17313868613138686,
+      "grad_norm": 1.4876685295483647,
+      "learning_rate": 9.472712631524599e-06,
+      "loss": 0.5423,
+      "step": 1779
+    },
+    {
+      "epoch": 0.17323600973236009,
+      "grad_norm": 1.4134646674679987,
+      "learning_rate": 9.472007894939624e-06,
+      "loss": 0.448,
+      "step": 1780
+    },
+    {
+      "epoch": 0.17333333333333334,
+      "grad_norm": 1.3805564537318322,
+      "learning_rate": 9.471302713968473e-06,
+      "loss": 0.2429,
+      "step": 1781
+    },
+    {
+      "epoch": 0.17343065693430657,
+      "grad_norm": 1.4256414475552066,
+      "learning_rate": 9.470597088681225e-06,
+      "loss": 0.4821,
+      "step": 1782
+    },
+    {
+      "epoch": 0.17352798053527982,
+      "grad_norm": 1.2857804565204727,
+      "learning_rate": 9.469891019147996e-06,
+      "loss": 0.3177,
+      "step": 1783
+    },
+    {
+      "epoch": 0.17362530413625304,
+      "grad_norm": 1.7384422656290006,
+      "learning_rate": 9.46918450543895e-06,
+      "loss": 0.6144,
+      "step": 1784
+    },
+    {
+      "epoch": 0.17372262773722627,
+      "grad_norm": 1.3733947226466707,
+      "learning_rate": 9.46847754762429e-06,
+      "loss": 0.3777,
+      "step": 1785
+    },
+    {
+      "epoch": 0.17381995133819952,
+      "grad_norm": 1.090627736959876,
+      "learning_rate": 9.467770145774271e-06,
+      "loss": 0.307,
+      "step": 1786
+    },
+    {
+      "epoch": 0.17391727493917275,
+      "grad_norm": 1.5306242617959314,
+      "learning_rate": 9.467062299959187e-06,
+      "loss": 0.4652,
+      "step": 1787
+    },
+    {
+      "epoch": 0.17401459854014598,
+      "grad_norm": 1.6335244702718128,
+      "learning_rate": 9.466354010249375e-06,
+      "loss": 0.5127,
+      "step": 1788
+    },
+    {
+      "epoch": 0.17411192214111923,
+      "grad_norm": 1.3582351114688258,
+      "learning_rate": 9.465645276715221e-06,
+      "loss": 0.4213,
+      "step": 1789
+    },
+    {
+      "epoch": 0.17420924574209246,
+      "grad_norm": 1.4962342995542501,
+      "learning_rate": 9.464936099427151e-06,
+      "loss": 0.4327,
+      "step": 1790
+    },
+    {
+      "epoch": 0.17430656934306568,
+      "grad_norm": 1.513533102257641,
+      "learning_rate": 9.464226478455636e-06,
+      "loss": 0.4527,
+      "step": 1791
+    },
+    {
+      "epoch": 0.17440389294403894,
+      "grad_norm": 1.4174664240767785,
+      "learning_rate": 9.463516413871193e-06,
+      "loss": 0.4986,
+      "step": 1792
+    },
+    {
+      "epoch": 0.17450121654501216,
+      "grad_norm": 1.283758777636687,
+      "learning_rate": 9.46280590574438e-06,
+      "loss": 0.4648,
+      "step": 1793
+    },
+    {
+      "epoch": 0.17459854014598541,
+      "grad_norm": 1.3960565511895506,
+      "learning_rate": 9.4620949541458e-06,
+      "loss": 0.3587,
+      "step": 1794
+    },
+    {
+      "epoch": 0.17469586374695864,
+      "grad_norm": 1.6199676647020385,
+      "learning_rate": 9.461383559146104e-06,
+      "loss": 0.5292,
+      "step": 1795
+    },
+    {
+      "epoch": 0.17479318734793187,
+      "grad_norm": 1.5028051531717803,
+      "learning_rate": 9.46067172081598e-06,
+      "loss": 0.4903,
+      "step": 1796
+    },
+    {
+      "epoch": 0.17489051094890512,
+      "grad_norm": 1.683063467822515,
+      "learning_rate": 9.459959439226165e-06,
+      "loss": 0.3106,
+      "step": 1797
+    },
+    {
+      "epoch": 0.17498783454987835,
+      "grad_norm": 1.3296224342860092,
+      "learning_rate": 9.459246714447439e-06,
+      "loss": 0.409,
+      "step": 1798
+    },
+    {
+      "epoch": 0.17508515815085157,
+      "grad_norm": 1.5847151231323486,
+      "learning_rate": 9.458533546550628e-06,
+      "loss": 0.4169,
+      "step": 1799
+    },
+    {
+      "epoch": 0.17518248175182483,
+      "grad_norm": 1.495253204796384,
+      "learning_rate": 9.457819935606596e-06,
+      "loss": 0.3753,
+      "step": 1800
+    },
+    {
+      "epoch": 0.17527980535279805,
+      "grad_norm": 1.4723876384358174,
+      "learning_rate": 9.45710588168626e-06,
+      "loss": 0.2437,
+      "step": 1801
+    },
+    {
+      "epoch": 0.17537712895377128,
+      "grad_norm": 1.5610295815557715,
+      "learning_rate": 9.45639138486057e-06,
+      "loss": 0.5651,
+      "step": 1802
+    },
+    {
+      "epoch": 0.17547445255474453,
+      "grad_norm": 1.5876154549734276,
+      "learning_rate": 9.45567644520053e-06,
+      "loss": 0.4835,
+      "step": 1803
+    },
+    {
+      "epoch": 0.17557177615571776,
+      "grad_norm": 1.5619759252942187,
+      "learning_rate": 9.454961062777181e-06,
+      "loss": 0.3036,
+      "step": 1804
+    },
+    {
+      "epoch": 0.175669099756691,
+      "grad_norm": 1.0144029160546408,
+      "learning_rate": 9.454245237661617e-06,
+      "loss": 0.219,
+      "step": 1805
+    },
+    {
+      "epoch": 0.17576642335766424,
+      "grad_norm": 1.717922774563162,
+      "learning_rate": 9.453528969924963e-06,
+      "loss": 0.5388,
+      "step": 1806
+    },
+    {
+      "epoch": 0.17586374695863746,
+      "grad_norm": 1.288743598100688,
+      "learning_rate": 9.452812259638399e-06,
+      "loss": 0.4171,
+      "step": 1807
+    },
+    {
+      "epoch": 0.17596107055961072,
+      "grad_norm": 1.951279890184611,
+      "learning_rate": 9.452095106873142e-06,
+      "loss": 0.3823,
+      "step": 1808
+    },
+    {
+      "epoch": 0.17605839416058394,
+      "grad_norm": 1.352467757455935,
+      "learning_rate": 9.45137751170046e-06,
+      "loss": 0.3137,
+      "step": 1809
+    },
+    {
+      "epoch": 0.17615571776155717,
+      "grad_norm": 1.3883395327139227,
+      "learning_rate": 9.450659474191658e-06,
+      "loss": 0.4878,
+      "step": 1810
+    },
+    {
+      "epoch": 0.17625304136253042,
+      "grad_norm": 1.5658708451700805,
+      "learning_rate": 9.449940994418088e-06,
+      "loss": 0.523,
+      "step": 1811
+    },
+    {
+      "epoch": 0.17635036496350365,
+      "grad_norm": 1.215080164631292,
+      "learning_rate": 9.449222072451147e-06,
+      "loss": 0.3773,
+      "step": 1812
+    },
+    {
+      "epoch": 0.17644768856447687,
+      "grad_norm": 1.524331324578441,
+      "learning_rate": 9.448502708362273e-06,
+      "loss": 0.539,
+      "step": 1813
+    },
+    {
+      "epoch": 0.17654501216545013,
+      "grad_norm": 1.6985132616371517,
+      "learning_rate": 9.447782902222951e-06,
+      "loss": 0.6344,
+      "step": 1814
+    },
+    {
+      "epoch": 0.17664233576642335,
+      "grad_norm": 1.394156226586294,
+      "learning_rate": 9.447062654104708e-06,
+      "loss": 0.4136,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1767396593673966,
+      "grad_norm": 1.0359913462457855,
+      "learning_rate": 9.446341964079116e-06,
+      "loss": 0.2471,
+      "step": 1816
+    },
+    {
+      "epoch": 0.17683698296836983,
+      "grad_norm": 1.6379291001324041,
+      "learning_rate": 9.44562083221779e-06,
+      "loss": 0.4648,
+      "step": 1817
+    },
+    {
+      "epoch": 0.17693430656934306,
+      "grad_norm": 1.0926982727654353,
+      "learning_rate": 9.44489925859239e-06,
+      "loss": 0.253,
+      "step": 1818
+    },
+    {
+      "epoch": 0.1770316301703163,
+      "grad_norm": 1.3396314447206463,
+      "learning_rate": 9.444177243274619e-06,
+      "loss": 0.4053,
+      "step": 1819
+    },
+    {
+      "epoch": 0.17712895377128954,
+      "grad_norm": 1.2170984864894128,
+      "learning_rate": 9.44345478633622e-06,
+      "loss": 0.3483,
+      "step": 1820
+    },
+    {
+      "epoch": 0.17722627737226276,
+      "grad_norm": 1.9241463489982464,
+      "learning_rate": 9.442731887848993e-06,
+      "loss": 0.7875,
+      "step": 1821
+    },
+    {
+      "epoch": 0.17732360097323602,
+      "grad_norm": 1.7367037011857493,
+      "learning_rate": 9.442008547884765e-06,
+      "loss": 0.5423,
+      "step": 1822
+    },
+    {
+      "epoch": 0.17742092457420924,
+      "grad_norm": 1.7768925691501514,
+      "learning_rate": 9.441284766515417e-06,
+      "loss": 0.5332,
+      "step": 1823
+    },
+    {
+      "epoch": 0.17751824817518247,
+      "grad_norm": 1.544872490519166,
+      "learning_rate": 9.440560543812872e-06,
+      "loss": 0.4797,
+      "step": 1824
+    },
+    {
+      "epoch": 0.17761557177615572,
+      "grad_norm": 1.3959412272112985,
+      "learning_rate": 9.439835879849097e-06,
+      "loss": 0.2813,
+      "step": 1825
+    },
+    {
+      "epoch": 0.17771289537712895,
+      "grad_norm": 1.4333698815114406,
+      "learning_rate": 9.439110774696101e-06,
+      "loss": 0.4623,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1778102189781022,
+      "grad_norm": 1.4483549520432324,
+      "learning_rate": 9.43838522842594e-06,
+      "loss": 0.3718,
+      "step": 1827
+    },
+    {
+      "epoch": 0.17790754257907543,
+      "grad_norm": 1.1321375447475677,
+      "learning_rate": 9.43765924111071e-06,
+      "loss": 0.3035,
+      "step": 1828
+    },
+    {
+      "epoch": 0.17800486618004865,
+      "grad_norm": 1.362326738732822,
+      "learning_rate": 9.436932812822554e-06,
+      "loss": 0.316,
+      "step": 1829
+    },
+    {
+      "epoch": 0.1781021897810219,
+      "grad_norm": 1.460799021966237,
+      "learning_rate": 9.436205943633656e-06,
+      "loss": 0.3911,
+      "step": 1830
+    },
+    {
+      "epoch": 0.17819951338199513,
+      "grad_norm": 1.5389161016090995,
+      "learning_rate": 9.435478633616247e-06,
+      "loss": 0.5521,
+      "step": 1831
+    },
+    {
+      "epoch": 0.17829683698296836,
+      "grad_norm": 1.5219868331018827,
+      "learning_rate": 9.4347508828426e-06,
+      "loss": 0.5027,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1783941605839416,
+      "grad_norm": 1.245087028586955,
+      "learning_rate": 9.434022691385034e-06,
+      "loss": 0.2981,
+      "step": 1833
+    },
+    {
+      "epoch": 0.17849148418491484,
+      "grad_norm": 1.4557548434245557,
+      "learning_rate": 9.433294059315905e-06,
+      "loss": 0.2293,
+      "step": 1834
+    },
+    {
+      "epoch": 0.17858880778588807,
+      "grad_norm": 1.3081558633618169,
+      "learning_rate": 9.432564986707621e-06,
+      "loss": 0.4217,
+      "step": 1835
+    },
+    {
+      "epoch": 0.17868613138686132,
+      "grad_norm": 1.3513560054673133,
+      "learning_rate": 9.43183547363263e-06,
+      "loss": 0.4318,
+      "step": 1836
+    },
+    {
+      "epoch": 0.17878345498783454,
+      "grad_norm": 1.3315264956466353,
+      "learning_rate": 9.431105520163426e-06,
+      "loss": 0.3781,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1788807785888078,
+      "grad_norm": 1.0550787306059675,
+      "learning_rate": 9.430375126372542e-06,
+      "loss": 0.3104,
+      "step": 1838
+    },
+    {
+      "epoch": 0.17897810218978102,
+      "grad_norm": 1.3337629142786684,
+      "learning_rate": 9.429644292332557e-06,
+      "loss": 0.3455,
+      "step": 1839
+    },
+    {
+      "epoch": 0.17907542579075425,
+      "grad_norm": 1.6239197882024916,
+      "learning_rate": 9.428913018116098e-06,
+      "loss": 0.5855,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1791727493917275,
+      "grad_norm": 1.3780162846249417,
+      "learning_rate": 9.428181303795828e-06,
+      "loss": 0.3643,
+      "step": 1841
+    },
+    {
+      "epoch": 0.17927007299270073,
+      "grad_norm": 1.3478310292007554,
+      "learning_rate": 9.42744914944446e-06,
+      "loss": 0.3962,
+      "step": 1842
+    },
+    {
+      "epoch": 0.17936739659367396,
+      "grad_norm": 1.5440243743593307,
+      "learning_rate": 9.426716555134751e-06,
+      "loss": 0.6193,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1794647201946472,
+      "grad_norm": 1.4878960058265709,
+      "learning_rate": 9.425983520939495e-06,
+      "loss": 0.473,
+      "step": 1844
+    },
+    {
+      "epoch": 0.17956204379562044,
+      "grad_norm": 1.672460221871015,
+      "learning_rate": 9.425250046931539e-06,
+      "loss": 0.6429,
+      "step": 1845
+    },
+    {
+      "epoch": 0.17965936739659366,
+      "grad_norm": 1.6015212635221012,
+      "learning_rate": 9.424516133183762e-06,
+      "loss": 0.3195,
+      "step": 1846
+    },
+    {
+      "epoch": 0.17975669099756691,
+      "grad_norm": 1.385761715171386,
+      "learning_rate": 9.4237817797691e-06,
+      "loss": 0.4054,
+      "step": 1847
+    },
+    {
+      "epoch": 0.17985401459854014,
+      "grad_norm": 1.386847906411032,
+      "learning_rate": 9.423046986760522e-06,
+      "loss": 0.3825,
+      "step": 1848
+    },
+    {
+      "epoch": 0.1799513381995134,
+      "grad_norm": 1.087510047515406,
+      "learning_rate": 9.422311754231047e-06,
+      "loss": 0.3213,
+      "step": 1849
+    },
+    {
+      "epoch": 0.18004866180048662,
+      "grad_norm": 1.6065416301387576,
+      "learning_rate": 9.421576082253734e-06,
+      "loss": 0.5062,
+      "step": 1850
+    },
+    {
+      "epoch": 0.18014598540145985,
+      "grad_norm": 1.34096451308299,
+      "learning_rate": 9.42083997090169e-06,
+      "loss": 0.4036,
+      "step": 1851
+    },
+    {
+      "epoch": 0.1802433090024331,
+      "grad_norm": 1.2557739418598393,
+      "learning_rate": 9.42010342024806e-06,
+      "loss": 0.3595,
+      "step": 1852
+    },
+    {
+      "epoch": 0.18034063260340633,
+      "grad_norm": 1.5281441778996137,
+      "learning_rate": 9.419366430366035e-06,
+      "loss": 0.604,
+      "step": 1853
+    },
+    {
+      "epoch": 0.18043795620437955,
+      "grad_norm": 1.2665309724570952,
+      "learning_rate": 9.418629001328852e-06,
+      "loss": 0.4205,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1805352798053528,
+      "grad_norm": 1.3442942382162348,
+      "learning_rate": 9.417891133209789e-06,
+      "loss": 0.3457,
+      "step": 1855
+    },
+    {
+      "epoch": 0.18063260340632603,
+      "grad_norm": 1.4106593198915445,
+      "learning_rate": 9.417152826082169e-06,
+      "loss": 0.4812,
+      "step": 1856
+    },
+    {
+      "epoch": 0.18072992700729926,
+      "grad_norm": 1.4377180846268287,
+      "learning_rate": 9.416414080019359e-06,
+      "loss": 0.4618,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1808272506082725,
+      "grad_norm": 3.1493721230250182,
+      "learning_rate": 9.415674895094765e-06,
+      "loss": 0.4636,
+      "step": 1858
+    },
+    {
+      "epoch": 0.18092457420924574,
+      "grad_norm": 1.2019926414899231,
+      "learning_rate": 9.414935271381844e-06,
+      "loss": 0.3081,
+      "step": 1859
+    },
+    {
+      "epoch": 0.181021897810219,
+      "grad_norm": 2.6470194483303042,
+      "learning_rate": 9.41419520895409e-06,
+      "loss": 0.545,
+      "step": 1860
+    },
+    {
+      "epoch": 0.18111922141119222,
+      "grad_norm": 1.2980614715591199,
+      "learning_rate": 9.413454707885048e-06,
+      "loss": 0.2964,
+      "step": 1861
+    },
+    {
+      "epoch": 0.18121654501216544,
+      "grad_norm": 1.0776172492719038,
+      "learning_rate": 9.412713768248296e-06,
+      "loss": 0.3014,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1813138686131387,
+      "grad_norm": 1.6105644497131084,
+      "learning_rate": 9.411972390117466e-06,
+      "loss": 0.2939,
+      "step": 1863
+    },
+    {
+      "epoch": 0.18141119221411192,
+      "grad_norm": 1.5656908641978677,
+      "learning_rate": 9.411230573566227e-06,
+      "loss": 0.5202,
+      "step": 1864
+    },
+    {
+      "epoch": 0.18150851581508515,
+      "grad_norm": 1.303806212869287,
+      "learning_rate": 9.410488318668294e-06,
+      "loss": 0.333,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1816058394160584,
+      "grad_norm": 1.6655746538236336,
+      "learning_rate": 9.409745625497427e-06,
+      "loss": 0.432,
+      "step": 1866
+    },
+    {
+      "epoch": 0.18170316301703163,
+      "grad_norm": 1.3843667729738216,
+      "learning_rate": 9.409002494127427e-06,
+      "loss": 0.3721,
+      "step": 1867
+    },
+    {
+      "epoch": 0.18180048661800485,
+      "grad_norm": 1.119511993732411,
+      "learning_rate": 9.408258924632139e-06,
+      "loss": 0.3344,
+      "step": 1868
+    },
+    {
+      "epoch": 0.1818978102189781,
+      "grad_norm": 1.402581324947916,
+      "learning_rate": 9.407514917085451e-06,
+      "loss": 0.4016,
+      "step": 1869
+    },
+    {
+      "epoch": 0.18199513381995133,
+      "grad_norm": 1.424239738841203,
+      "learning_rate": 9.406770471561298e-06,
+      "loss": 0.4043,
+      "step": 1870
+    },
+    {
+      "epoch": 0.18209245742092459,
+      "grad_norm": 1.4825401610777273,
+      "learning_rate": 9.406025588133654e-06,
+      "loss": 0.5446,
+      "step": 1871
+    },
+    {
+      "epoch": 0.1821897810218978,
+      "grad_norm": 1.1812973154269832,
+      "learning_rate": 9.405280266876539e-06,
+      "loss": 0.3086,
+      "step": 1872
+    },
+    {
+      "epoch": 0.18228710462287104,
+      "grad_norm": 1.458454653825207,
+      "learning_rate": 9.404534507864015e-06,
+      "loss": 0.426,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1823844282238443,
+      "grad_norm": 1.4345175445802738,
+      "learning_rate": 9.403788311170193e-06,
+      "loss": 0.4826,
+      "step": 1874
+    },
+    {
+      "epoch": 0.18248175182481752,
+      "grad_norm": 1.636664123351898,
+      "learning_rate": 9.403041676869217e-06,
+      "loss": 0.5861,
+      "step": 1875
+    },
+    {
+      "epoch": 0.18257907542579074,
+      "grad_norm": 1.4112207510715695,
+      "learning_rate": 9.402294605035285e-06,
+      "loss": 0.3575,
+      "step": 1876
+    },
+    {
+      "epoch": 0.182676399026764,
+      "grad_norm": 1.5632317164864975,
+      "learning_rate": 9.401547095742631e-06,
+      "loss": 0.5798,
+      "step": 1877
+    },
+    {
+      "epoch": 0.18277372262773722,
+      "grad_norm": 1.2700759423445944,
+      "learning_rate": 9.400799149065538e-06,
+      "loss": 0.3928,
+      "step": 1878
+    },
+    {
+      "epoch": 0.18287104622871045,
+      "grad_norm": 1.1318646905388465,
+      "learning_rate": 9.400050765078327e-06,
+      "loss": 0.2783,
+      "step": 1879
+    },
+    {
+      "epoch": 0.1829683698296837,
+      "grad_norm": 1.1697084872304198,
+      "learning_rate": 9.399301943855368e-06,
+      "loss": 0.2715,
+      "step": 1880
+    },
+    {
+      "epoch": 0.18306569343065693,
+      "grad_norm": 1.4137887426273796,
+      "learning_rate": 9.39855268547107e-06,
+      "loss": 0.3049,
+      "step": 1881
+    },
+    {
+      "epoch": 0.18316301703163018,
+      "grad_norm": 1.3869164554267486,
+      "learning_rate": 9.397802989999888e-06,
+      "loss": 0.3526,
+      "step": 1882
+    },
+    {
+      "epoch": 0.1832603406326034,
+      "grad_norm": 1.3336674996684654,
+      "learning_rate": 9.39705285751632e-06,
+      "loss": 0.3914,
+      "step": 1883
+    },
+    {
+      "epoch": 0.18335766423357663,
+      "grad_norm": 1.2095628873380657,
+      "learning_rate": 9.396302288094907e-06,
+      "loss": 0.3577,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1834549878345499,
+      "grad_norm": 1.4741118747641506,
+      "learning_rate": 9.395551281810233e-06,
+      "loss": 0.4753,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1835523114355231,
+      "grad_norm": 1.5440799623052803,
+      "learning_rate": 9.394799838736928e-06,
+      "loss": 0.5143,
+      "step": 1886
+    },
+    {
+      "epoch": 0.18364963503649634,
+      "grad_norm": 1.6461828641301555,
+      "learning_rate": 9.394047958949661e-06,
+      "loss": 0.5046,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1837469586374696,
+      "grad_norm": 1.3077272649446732,
+      "learning_rate": 9.393295642523147e-06,
+      "loss": 0.4505,
+      "step": 1888
+    },
+    {
+      "epoch": 0.18384428223844282,
+      "grad_norm": 1.3954964938282017,
+      "learning_rate": 9.392542889532146e-06,
+      "loss": 0.3752,
+      "step": 1889
+    },
+    {
+      "epoch": 0.18394160583941604,
+      "grad_norm": 1.4332674159188397,
+      "learning_rate": 9.391789700051457e-06,
+      "loss": 0.4102,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1840389294403893,
+      "grad_norm": 1.5291760471205262,
+      "learning_rate": 9.391036074155926e-06,
+      "loss": 0.3892,
+      "step": 1891
+    },
+    {
+      "epoch": 0.18413625304136252,
+      "grad_norm": 1.3194046059109847,
+      "learning_rate": 9.390282011920442e-06,
+      "loss": 0.3402,
+      "step": 1892
+    },
+    {
+      "epoch": 0.18423357664233578,
+      "grad_norm": 1.1218553674196712,
+      "learning_rate": 9.389527513419935e-06,
+      "loss": 0.2705,
+      "step": 1893
+    },
+    {
+      "epoch": 0.184330900243309,
+      "grad_norm": 1.4415924034763155,
+      "learning_rate": 9.388772578729382e-06,
+      "loss": 0.4153,
+      "step": 1894
+    },
+    {
+      "epoch": 0.18442822384428223,
+      "grad_norm": 1.1449469634853555,
+      "learning_rate": 9.3880172079238e-06,
+      "loss": 0.2464,
+      "step": 1895
+    },
+    {
+      "epoch": 0.18452554744525548,
+      "grad_norm": 1.3609647553229742,
+      "learning_rate": 9.38726140107825e-06,
+      "loss": 0.4167,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1846228710462287,
+      "grad_norm": 1.5005607351629322,
+      "learning_rate": 9.38650515826784e-06,
+      "loss": 0.5496,
+      "step": 1897
+    },
+    {
+      "epoch": 0.18472019464720194,
+      "grad_norm": 1.2988771816540412,
+      "learning_rate": 9.385748479567715e-06,
+      "loss": 0.3746,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1848175182481752,
+      "grad_norm": 1.6297457427665438,
+      "learning_rate": 9.384991365053066e-06,
+      "loss": 0.5329,
+      "step": 1899
+    },
+    {
+      "epoch": 0.18491484184914841,
+      "grad_norm": 1.4260746902123356,
+      "learning_rate": 9.384233814799133e-06,
+      "loss": 0.5495,
+      "step": 1900
+    },
+    {
+      "epoch": 0.18501216545012167,
+      "grad_norm": 1.6131616876000299,
+      "learning_rate": 9.38347582888119e-06,
+      "loss": 0.4956,
+      "step": 1901
+    },
+    {
+      "epoch": 0.1851094890510949,
+      "grad_norm": 1.2427047036633028,
+      "learning_rate": 9.382717407374559e-06,
+      "loss": 0.3527,
+      "step": 1902
+    },
+    {
+      "epoch": 0.18520681265206812,
+      "grad_norm": 1.1650358905093554,
+      "learning_rate": 9.381958550354607e-06,
+      "loss": 0.3282,
+      "step": 1903
+    },
+    {
+      "epoch": 0.18530413625304137,
+      "grad_norm": 1.2422827918011654,
+      "learning_rate": 9.381199257896738e-06,
+      "loss": 0.3954,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1854014598540146,
+      "grad_norm": 1.3772059864511268,
+      "learning_rate": 9.38043953007641e-06,
+      "loss": 0.2519,
+      "step": 1905
+    },
+    {
+      "epoch": 0.18549878345498783,
+      "grad_norm": 1.2627132972091453,
+      "learning_rate": 9.379679366969108e-06,
+      "loss": 0.3748,
+      "step": 1906
+    },
+    {
+      "epoch": 0.18559610705596108,
+      "grad_norm": 1.7742544300786764,
+      "learning_rate": 9.378918768650379e-06,
+      "loss": 0.4627,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1856934306569343,
+      "grad_norm": 1.3460661864821146,
+      "learning_rate": 9.3781577351958e-06,
+      "loss": 0.3769,
+      "step": 1908
+    },
+    {
+      "epoch": 0.18579075425790753,
+      "grad_norm": 1.2948363493096455,
+      "learning_rate": 9.377396266680993e-06,
+      "loss": 0.255,
+      "step": 1909
+    },
+    {
+      "epoch": 0.18588807785888078,
+      "grad_norm": 1.4260435934265066,
+      "learning_rate": 9.376634363181631e-06,
+      "loss": 0.4158,
+      "step": 1910
+    },
+    {
+      "epoch": 0.185985401459854,
+      "grad_norm": 1.4136193355548345,
+      "learning_rate": 9.375872024773423e-06,
+      "loss": 0.3764,
+      "step": 1911
+    },
+    {
+      "epoch": 0.18608272506082726,
+      "grad_norm": 1.2338333390059972,
+      "learning_rate": 9.375109251532121e-06,
+      "loss": 0.3785,
+      "step": 1912
+    },
+    {
+      "epoch": 0.1861800486618005,
+      "grad_norm": 1.535249430616727,
+      "learning_rate": 9.374346043533524e-06,
+      "loss": 0.5252,
+      "step": 1913
+    },
+    {
+      "epoch": 0.18627737226277372,
+      "grad_norm": 1.215284604855692,
+      "learning_rate": 9.373582400853472e-06,
+      "loss": 0.3295,
+      "step": 1914
+    },
+    {
+      "epoch": 0.18637469586374697,
+      "grad_norm": 1.331605367733698,
+      "learning_rate": 9.372818323567847e-06,
+      "loss": 0.2818,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1864720194647202,
+      "grad_norm": 1.3700650260278666,
+      "learning_rate": 9.37205381175258e-06,
+      "loss": 0.5125,
+      "step": 1916
+    },
+    {
+      "epoch": 0.18656934306569342,
+      "grad_norm": 1.0730618437287824,
+      "learning_rate": 9.371288865483637e-06,
+      "loss": 0.3608,
+      "step": 1917
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 1.6775147335354874,
+      "learning_rate": 9.370523484837033e-06,
+      "loss": 0.4555,
+      "step": 1918
+    },
+    {
+      "epoch": 0.1867639902676399,
+      "grad_norm": 1.531630799569193,
+      "learning_rate": 9.369757669888822e-06,
+      "loss": 0.502,
+      "step": 1919
+    },
+    {
+      "epoch": 0.18686131386861313,
+      "grad_norm": 0.924734272033398,
+      "learning_rate": 9.368991420715109e-06,
+      "loss": 0.2117,
+      "step": 1920
+    },
+    {
+      "epoch": 0.18695863746958638,
+      "grad_norm": 1.3568146369682141,
+      "learning_rate": 9.36822473739203e-06,
+      "loss": 0.4311,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1870559610705596,
+      "grad_norm": 1.2577909858711795,
+      "learning_rate": 9.367457619995776e-06,
+      "loss": 0.405,
+      "step": 1922
+    },
+    {
+      "epoch": 0.18715328467153286,
+      "grad_norm": 1.5933524739274278,
+      "learning_rate": 9.366690068602573e-06,
+      "loss": 0.627,
+      "step": 1923
+    },
+    {
+      "epoch": 0.18725060827250609,
+      "grad_norm": 1.279419778059805,
+      "learning_rate": 9.365922083288694e-06,
+      "loss": 0.2814,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1873479318734793,
+      "grad_norm": 1.6336124778487715,
+      "learning_rate": 9.365153664130454e-06,
+      "loss": 0.6461,
+      "step": 1925
+    },
+    {
+      "epoch": 0.18744525547445257,
+      "grad_norm": 5.906434394339674,
+      "learning_rate": 9.364384811204212e-06,
+      "loss": 0.5628,
+      "step": 1926
+    },
+    {
+      "epoch": 0.1875425790754258,
+      "grad_norm": 1.2770793302804129,
+      "learning_rate": 9.363615524586368e-06,
+      "loss": 0.303,
+      "step": 1927
+    },
+    {
+      "epoch": 0.18763990267639902,
+      "grad_norm": 1.2695156624644028,
+      "learning_rate": 9.362845804353367e-06,
+      "loss": 0.3592,
+      "step": 1928
+    },
+    {
+      "epoch": 0.18773722627737227,
+      "grad_norm": 1.4443375056776053,
+      "learning_rate": 9.362075650581698e-06,
+      "loss": 0.4701,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1878345498783455,
+      "grad_norm": 1.4330727776563095,
+      "learning_rate": 9.36130506334789e-06,
+      "loss": 0.5163,
+      "step": 1930
+    },
+    {
+      "epoch": 0.18793187347931872,
+      "grad_norm": 1.326934280688427,
+      "learning_rate": 9.360534042728517e-06,
+      "loss": 0.289,
+      "step": 1931
+    },
+    {
+      "epoch": 0.18802919708029198,
+      "grad_norm": 1.0531370847104877,
+      "learning_rate": 9.359762588800195e-06,
+      "loss": 0.1994,
+      "step": 1932
+    },
+    {
+      "epoch": 0.1881265206812652,
+      "grad_norm": 1.4998435892573359,
+      "learning_rate": 9.358990701639585e-06,
+      "loss": 0.4064,
+      "step": 1933
+    },
+    {
+      "epoch": 0.18822384428223846,
+      "grad_norm": 2.65155925581941,
+      "learning_rate": 9.358218381323391e-06,
+      "loss": 0.3513,
+      "step": 1934
+    },
+    {
+      "epoch": 0.18832116788321168,
+      "grad_norm": 1.280523326704506,
+      "learning_rate": 9.357445627928356e-06,
+      "loss": 0.3132,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1884184914841849,
+      "grad_norm": 1.347047087613105,
+      "learning_rate": 9.356672441531273e-06,
+      "loss": 0.3334,
+      "step": 1936
+    },
+    {
+      "epoch": 0.18851581508515816,
+      "grad_norm": 1.2987558904079175,
+      "learning_rate": 9.35589882220897e-06,
+      "loss": 0.3224,
+      "step": 1937
+    },
+    {
+      "epoch": 0.1886131386861314,
+      "grad_norm": 0.9974048438134153,
+      "learning_rate": 9.355124770038323e-06,
+      "loss": 0.2764,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1887104622871046,
+      "grad_norm": 2.544180913694316,
+      "learning_rate": 9.354350285096255e-06,
+      "loss": 0.495,
+      "step": 1939
+    },
+    {
+      "epoch": 0.18880778588807787,
+      "grad_norm": 1.613510595834776,
+      "learning_rate": 9.353575367459718e-06,
+      "loss": 0.5269,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1889051094890511,
+      "grad_norm": 1.1663508101189002,
+      "learning_rate": 9.352800017205724e-06,
+      "loss": 0.3936,
+      "step": 1941
+    },
+    {
+      "epoch": 0.18900243309002432,
+      "grad_norm": 1.3673811421181858,
+      "learning_rate": 9.352024234411315e-06,
+      "loss": 0.4448,
+      "step": 1942
+    },
+    {
+      "epoch": 0.18909975669099757,
+      "grad_norm": 1.1481373712644614,
+      "learning_rate": 9.351248019153582e-06,
+      "loss": 0.3226,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1891970802919708,
+      "grad_norm": 1.025014870233366,
+      "learning_rate": 9.350471371509659e-06,
+      "loss": 0.2095,
+      "step": 1944
+    },
+    {
+      "epoch": 0.18929440389294405,
+      "grad_norm": 1.6587902238420225,
+      "learning_rate": 9.349694291556723e-06,
+      "loss": 0.3805,
+      "step": 1945
+    },
+    {
+      "epoch": 0.18939172749391728,
+      "grad_norm": 1.568770301353131,
+      "learning_rate": 9.348916779371993e-06,
+      "loss": 0.3902,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1894890510948905,
+      "grad_norm": 1.4274566422005779,
+      "learning_rate": 9.348138835032727e-06,
+      "loss": 0.3644,
+      "step": 1947
+    },
+    {
+      "epoch": 0.18958637469586376,
+      "grad_norm": 1.6590398647584288,
+      "learning_rate": 9.347360458616233e-06,
+      "loss": 0.3522,
+      "step": 1948
+    },
+    {
+      "epoch": 0.18968369829683698,
+      "grad_norm": 1.5905934658559544,
+      "learning_rate": 9.346581650199859e-06,
+      "loss": 0.3784,
+      "step": 1949
+    },
+    {
+      "epoch": 0.1897810218978102,
+      "grad_norm": 1.358850838464726,
+      "learning_rate": 9.345802409860995e-06,
+      "loss": 0.3407,
+      "step": 1950
+    },
+    {
+      "epoch": 0.18987834549878346,
+      "grad_norm": 1.5906740312195304,
+      "learning_rate": 9.345022737677073e-06,
+      "loss": 0.4735,
+      "step": 1951
+    },
+    {
+      "epoch": 0.1899756690997567,
+      "grad_norm": 1.419279223309371,
+      "learning_rate": 9.344242633725573e-06,
+      "loss": 0.4677,
+      "step": 1952
+    },
+    {
+      "epoch": 0.19007299270072991,
+      "grad_norm": 2.368125402390624,
+      "learning_rate": 9.34346209808401e-06,
+      "loss": 0.4341,
+      "step": 1953
+    },
+    {
+      "epoch": 0.19017031630170317,
+      "grad_norm": 1.6018933954570558,
+      "learning_rate": 9.342681130829949e-06,
+      "loss": 0.4348,
+      "step": 1954
+    },
+    {
+      "epoch": 0.1902676399026764,
+      "grad_norm": 1.4757982324740848,
+      "learning_rate": 9.341899732040996e-06,
+      "loss": 0.393,
+      "step": 1955
+    },
+    {
+      "epoch": 0.19036496350364965,
+      "grad_norm": 1.463093762624457,
+      "learning_rate": 9.341117901794797e-06,
+      "loss": 0.3787,
+      "step": 1956
+    },
+    {
+      "epoch": 0.19046228710462287,
+      "grad_norm": 1.5507561900230402,
+      "learning_rate": 9.340335640169045e-06,
+      "loss": 0.4715,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1905596107055961,
+      "grad_norm": 1.4207468273121375,
+      "learning_rate": 9.339552947241471e-06,
+      "loss": 0.3938,
+      "step": 1958
+    },
+    {
+      "epoch": 0.19065693430656935,
+      "grad_norm": 1.407596113402629,
+      "learning_rate": 9.338769823089853e-06,
+      "loss": 0.4965,
+      "step": 1959
+    },
+    {
+      "epoch": 0.19075425790754258,
+      "grad_norm": 1.5505869092648736,
+      "learning_rate": 9.337986267792014e-06,
+      "loss": 0.3699,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1908515815085158,
+      "grad_norm": 1.4558635051434323,
+      "learning_rate": 9.33720228142581e-06,
+      "loss": 0.3436,
+      "step": 1961
+    },
+    {
+      "epoch": 0.19094890510948906,
+      "grad_norm": 1.4210127007858437,
+      "learning_rate": 9.336417864069152e-06,
+      "loss": 0.3959,
+      "step": 1962
+    },
+    {
+      "epoch": 0.19104622871046228,
+      "grad_norm": 1.5797691467496429,
+      "learning_rate": 9.335633015799983e-06,
+      "loss": 0.5438,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1911435523114355,
+      "grad_norm": 1.200940613853037,
+      "learning_rate": 9.334847736696297e-06,
+      "loss": 0.3037,
+      "step": 1964
+    },
+    {
+      "epoch": 0.19124087591240876,
+      "grad_norm": 1.6206966051553,
+      "learning_rate": 9.334062026836128e-06,
+      "loss": 0.6412,
+      "step": 1965
+    },
+    {
+      "epoch": 0.191338199513382,
+      "grad_norm": 1.3678147539203456,
+      "learning_rate": 9.33327588629755e-06,
+      "loss": 0.328,
+      "step": 1966
+    },
+    {
+      "epoch": 0.19143552311435524,
+      "grad_norm": 1.425436568728509,
+      "learning_rate": 9.332489315158685e-06,
+      "loss": 0.42,
+      "step": 1967
+    },
+    {
+      "epoch": 0.19153284671532847,
+      "grad_norm": 1.4740185495034979,
+      "learning_rate": 9.331702313497693e-06,
+      "loss": 0.3563,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1916301703163017,
+      "grad_norm": 1.4865130636524604,
+      "learning_rate": 9.33091488139278e-06,
+      "loss": 0.3452,
+      "step": 1969
+    },
+    {
+      "epoch": 0.19172749391727495,
+      "grad_norm": 1.595704917953399,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.6593,
+      "step": 1970
+    },
+    {
+      "epoch": 0.19182481751824818,
+      "grad_norm": 1.4305855687191487,
+      "learning_rate": 9.329338726164225e-06,
+      "loss": 0.4935,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1919221411192214,
+      "grad_norm": 1.4810316480182457,
+      "learning_rate": 9.328550003197203e-06,
+      "loss": 0.4303,
+      "step": 1972
+    },
+    {
+      "epoch": 0.19201946472019465,
+      "grad_norm": 1.1937939840472271,
+      "learning_rate": 9.32776085009951e-06,
+      "loss": 0.3178,
+      "step": 1973
+    },
+    {
+      "epoch": 0.19211678832116788,
+      "grad_norm": 1.3344201288029265,
+      "learning_rate": 9.326971266949558e-06,
+      "loss": 0.3469,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1922141119221411,
+      "grad_norm": 1.5818137690503504,
+      "learning_rate": 9.326181253825813e-06,
+      "loss": 0.505,
+      "step": 1975
+    },
+    {
+      "epoch": 0.19231143552311436,
+      "grad_norm": 1.263126969220317,
+      "learning_rate": 9.325390810806778e-06,
+      "loss": 0.3967,
+      "step": 1976
+    },
+    {
+      "epoch": 0.19240875912408759,
+      "grad_norm": 1.6967730581105949,
+      "learning_rate": 9.324599937971e-06,
+      "loss": 0.7353,
+      "step": 1977
+    },
+    {
+      "epoch": 0.19250608272506084,
+      "grad_norm": 1.4550804189369502,
+      "learning_rate": 9.323808635397067e-06,
+      "loss": 0.3326,
+      "step": 1978
+    },
+    {
+      "epoch": 0.19260340632603407,
+      "grad_norm": 1.594493767215082,
+      "learning_rate": 9.323016903163612e-06,
+      "loss": 0.4547,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1927007299270073,
+      "grad_norm": 1.4855552398261571,
+      "learning_rate": 9.322224741349313e-06,
+      "loss": 0.5095,
+      "step": 1980
+    },
+    {
+      "epoch": 0.19279805352798055,
+      "grad_norm": 1.3769945503658922,
+      "learning_rate": 9.321432150032884e-06,
+      "loss": 0.3853,
+      "step": 1981
+    },
+    {
+      "epoch": 0.19289537712895377,
+      "grad_norm": 1.3138128708042736,
+      "learning_rate": 9.320639129293083e-06,
+      "loss": 0.4129,
+      "step": 1982
+    },
+    {
+      "epoch": 0.192992700729927,
+      "grad_norm": 1.4617598559962484,
+      "learning_rate": 9.319845679208719e-06,
+      "loss": 0.449,
+      "step": 1983
+    },
+    {
+      "epoch": 0.19309002433090025,
+      "grad_norm": 1.6332060417216765,
+      "learning_rate": 9.319051799858633e-06,
+      "loss": 0.594,
+      "step": 1984
+    },
+    {
+      "epoch": 0.19318734793187348,
+      "grad_norm": 1.5432637765560855,
+      "learning_rate": 9.318257491321714e-06,
+      "loss": 0.3465,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1932846715328467,
+      "grad_norm": 1.4536395238750577,
+      "learning_rate": 9.317462753676895e-06,
+      "loss": 0.4212,
+      "step": 1986
+    },
+    {
+      "epoch": 0.19338199513381996,
+      "grad_norm": 1.3985266204226148,
+      "learning_rate": 9.31666758700315e-06,
+      "loss": 0.5313,
+      "step": 1987
+    },
+    {
+      "epoch": 0.19347931873479318,
+      "grad_norm": 1.4329939166816383,
+      "learning_rate": 9.315871991379493e-06,
+      "loss": 0.3958,
+      "step": 1988
+    },
+    {
+      "epoch": 0.19357664233576644,
+      "grad_norm": 1.3666417803863316,
+      "learning_rate": 9.315075966884984e-06,
+      "loss": 0.462,
+      "step": 1989
+    },
+    {
+      "epoch": 0.19367396593673966,
+      "grad_norm": 1.6059064802064114,
+      "learning_rate": 9.314279513598721e-06,
+      "loss": 0.5734,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1937712895377129,
+      "grad_norm": 1.521730062801285,
+      "learning_rate": 9.313482631599854e-06,
+      "loss": 0.3479,
+      "step": 1991
+    },
+    {
+      "epoch": 0.19386861313868614,
+      "grad_norm": 1.5212897395363751,
+      "learning_rate": 9.312685320967566e-06,
+      "loss": 0.4328,
+      "step": 1992
+    },
+    {
+      "epoch": 0.19396593673965937,
+      "grad_norm": 1.669365255826549,
+      "learning_rate": 9.311887581781086e-06,
+      "loss": 0.6153,
+      "step": 1993
+    },
+    {
+      "epoch": 0.1940632603406326,
+      "grad_norm": 1.1692329123053622,
+      "learning_rate": 9.311089414119688e-06,
+      "loss": 0.3149,
+      "step": 1994
+    },
+    {
+      "epoch": 0.19416058394160585,
+      "grad_norm": 1.4724909439197027,
+      "learning_rate": 9.310290818062683e-06,
+      "loss": 0.478,
+      "step": 1995
+    },
+    {
+      "epoch": 0.19425790754257907,
+      "grad_norm": 1.667688851021317,
+      "learning_rate": 9.309491793689431e-06,
+      "loss": 0.6192,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1943552311435523,
+      "grad_norm": 1.2423474670669281,
+      "learning_rate": 9.30869234107933e-06,
+      "loss": 0.4242,
+      "step": 1997
+    },
+    {
+      "epoch": 0.19445255474452555,
+      "grad_norm": 1.4117486896728357,
+      "learning_rate": 9.307892460311825e-06,
+      "loss": 0.4417,
+      "step": 1998
+    },
+    {
+      "epoch": 0.19454987834549878,
+      "grad_norm": 1.6605518542896853,
+      "learning_rate": 9.307092151466397e-06,
+      "loss": 0.5289,
+      "step": 1999
+    },
+    {
+      "epoch": 0.19464720194647203,
+      "grad_norm": 1.661933360658536,
+      "learning_rate": 9.306291414622575e-06,
+      "loss": 0.3357,
+      "step": 2000
+    },
+    {
+      "epoch": 0.19474452554744526,
+      "grad_norm": 1.4409618985011814,
+      "learning_rate": 9.305490249859927e-06,
+      "loss": 0.4563,
+      "step": 2001
+    },
+    {
+      "epoch": 0.19484184914841848,
+      "grad_norm": 1.9082899591217046,
+      "learning_rate": 9.304688657258068e-06,
+      "loss": 0.3445,
+      "step": 2002
+    },
+    {
+      "epoch": 0.19493917274939174,
+      "grad_norm": 1.2157434891172034,
+      "learning_rate": 9.303886636896649e-06,
+      "loss": 0.3719,
+      "step": 2003
+    },
+    {
+      "epoch": 0.19503649635036496,
+      "grad_norm": 1.57236888854409,
+      "learning_rate": 9.303084188855371e-06,
+      "loss": 0.4399,
+      "step": 2004
+    },
+    {
+      "epoch": 0.1951338199513382,
+      "grad_norm": 1.4041570559360463,
+      "learning_rate": 9.302281313213973e-06,
+      "loss": 0.4442,
+      "step": 2005
+    },
+    {
+      "epoch": 0.19523114355231144,
+      "grad_norm": 1.595081147428658,
+      "learning_rate": 9.301478010052237e-06,
+      "loss": 0.4225,
+      "step": 2006
+    },
+    {
+      "epoch": 0.19532846715328467,
+      "grad_norm": 1.562924823229517,
+      "learning_rate": 9.300674279449986e-06,
+      "loss": 0.3739,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1954257907542579,
+      "grad_norm": 1.6925679153497177,
+      "learning_rate": 9.299870121487088e-06,
+      "loss": 0.4465,
+      "step": 2008
+    },
+    {
+      "epoch": 0.19552311435523115,
+      "grad_norm": 1.4955175500348226,
+      "learning_rate": 9.299065536243453e-06,
+      "loss": 0.5055,
+      "step": 2009
+    },
+    {
+      "epoch": 0.19562043795620437,
+      "grad_norm": 1.5602814755448668,
+      "learning_rate": 9.298260523799035e-06,
+      "loss": 0.4214,
+      "step": 2010
+    },
+    {
+      "epoch": 0.19571776155717763,
+      "grad_norm": 1.4678189187481074,
+      "learning_rate": 9.297455084233826e-06,
+      "loss": 0.4221,
+      "step": 2011
+    },
+    {
+      "epoch": 0.19581508515815085,
+      "grad_norm": 1.1014848505883976,
+      "learning_rate": 9.296649217627863e-06,
+      "loss": 0.2531,
+      "step": 2012
+    },
+    {
+      "epoch": 0.19591240875912408,
+      "grad_norm": 1.553421501855423,
+      "learning_rate": 9.295842924061227e-06,
+      "loss": 0.5409,
+      "step": 2013
+    },
+    {
+      "epoch": 0.19600973236009733,
+      "grad_norm": 1.598118050761176,
+      "learning_rate": 9.295036203614039e-06,
+      "loss": 0.4084,
+      "step": 2014
+    },
+    {
+      "epoch": 0.19610705596107056,
+      "grad_norm": 1.6278848716274248,
+      "learning_rate": 9.294229056366464e-06,
+      "loss": 0.5842,
+      "step": 2015
+    },
+    {
+      "epoch": 0.19620437956204378,
+      "grad_norm": 1.243515264701947,
+      "learning_rate": 9.293421482398708e-06,
+      "loss": 0.3504,
+      "step": 2016
+    },
+    {
+      "epoch": 0.19630170316301704,
+      "grad_norm": 1.4687425329140307,
+      "learning_rate": 9.29261348179102e-06,
+      "loss": 0.2732,
+      "step": 2017
+    },
+    {
+      "epoch": 0.19639902676399026,
+      "grad_norm": 1.8000259635960119,
+      "learning_rate": 9.291805054623691e-06,
+      "loss": 0.7865,
+      "step": 2018
+    },
+    {
+      "epoch": 0.1964963503649635,
+      "grad_norm": 1.5721673591186547,
+      "learning_rate": 9.290996200977058e-06,
+      "loss": 0.5686,
+      "step": 2019
+    },
+    {
+      "epoch": 0.19659367396593674,
+      "grad_norm": 1.4634877349944297,
+      "learning_rate": 9.290186920931493e-06,
+      "loss": 0.4884,
+      "step": 2020
+    },
+    {
+      "epoch": 0.19669099756690997,
+      "grad_norm": 1.8795352763168436,
+      "learning_rate": 9.289377214567418e-06,
+      "loss": 0.279,
+      "step": 2021
+    },
+    {
+      "epoch": 0.19678832116788322,
+      "grad_norm": 1.2525962570268505,
+      "learning_rate": 9.288567081965292e-06,
+      "loss": 0.3003,
+      "step": 2022
+    },
+    {
+      "epoch": 0.19688564476885645,
+      "grad_norm": 1.4414518188882164,
+      "learning_rate": 9.28775652320562e-06,
+      "loss": 0.2883,
+      "step": 2023
+    },
+    {
+      "epoch": 0.19698296836982968,
+      "grad_norm": 1.1469869990322892,
+      "learning_rate": 9.286945538368946e-06,
+      "loss": 0.301,
+      "step": 2024
+    },
+    {
+      "epoch": 0.19708029197080293,
+      "grad_norm": 1.4386800814955665,
+      "learning_rate": 9.286134127535859e-06,
+      "loss": 0.417,
+      "step": 2025
+    },
+    {
+      "epoch": 0.19717761557177615,
+      "grad_norm": 1.4334168701816348,
+      "learning_rate": 9.28532229078699e-06,
+      "loss": 0.4694,
+      "step": 2026
+    },
+    {
+      "epoch": 0.19727493917274938,
+      "grad_norm": 1.2925159318336792,
+      "learning_rate": 9.28451002820301e-06,
+      "loss": 0.4438,
+      "step": 2027
+    },
+    {
+      "epoch": 0.19737226277372263,
+      "grad_norm": 1.1608723700468837,
+      "learning_rate": 9.283697339864635e-06,
+      "loss": 0.3899,
+      "step": 2028
+    },
+    {
+      "epoch": 0.19746958637469586,
+      "grad_norm": 1.0831308664734243,
+      "learning_rate": 9.282884225852625e-06,
+      "loss": 0.3594,
+      "step": 2029
+    },
+    {
+      "epoch": 0.19756690997566909,
+      "grad_norm": 1.3854325468066278,
+      "learning_rate": 9.282070686247773e-06,
+      "loss": 0.5111,
+      "step": 2030
+    },
+    {
+      "epoch": 0.19766423357664234,
+      "grad_norm": 1.2843702051671877,
+      "learning_rate": 9.281256721130927e-06,
+      "loss": 0.3298,
+      "step": 2031
+    },
+    {
+      "epoch": 0.19776155717761557,
+      "grad_norm": 1.4725158786403292,
+      "learning_rate": 9.280442330582968e-06,
+      "loss": 0.4776,
+      "step": 2032
+    },
+    {
+      "epoch": 0.19785888077858882,
+      "grad_norm": 1.2748346913452204,
+      "learning_rate": 9.279627514684826e-06,
+      "loss": 0.4438,
+      "step": 2033
+    },
+    {
+      "epoch": 0.19795620437956205,
+      "grad_norm": 1.406716290626126,
+      "learning_rate": 9.278812273517465e-06,
+      "loss": 0.2814,
+      "step": 2034
+    },
+    {
+      "epoch": 0.19805352798053527,
+      "grad_norm": 1.3303438388967537,
+      "learning_rate": 9.2779966071619e-06,
+      "loss": 0.4314,
+      "step": 2035
+    },
+    {
+      "epoch": 0.19815085158150852,
+      "grad_norm": 1.4134730169408085,
+      "learning_rate": 9.277180515699183e-06,
+      "loss": 0.2764,
+      "step": 2036
+    },
+    {
+      "epoch": 0.19824817518248175,
+      "grad_norm": 1.3255645305073551,
+      "learning_rate": 9.276363999210407e-06,
+      "loss": 0.4347,
+      "step": 2037
+    },
+    {
+      "epoch": 0.19834549878345498,
+      "grad_norm": 1.4369644328356708,
+      "learning_rate": 9.275547057776713e-06,
+      "loss": 0.3551,
+      "step": 2038
+    },
+    {
+      "epoch": 0.19844282238442823,
+      "grad_norm": 1.748281657046459,
+      "learning_rate": 9.27472969147928e-06,
+      "loss": 0.4372,
+      "step": 2039
+    },
+    {
+      "epoch": 0.19854014598540146,
+      "grad_norm": 1.2795189118800725,
+      "learning_rate": 9.273911900399331e-06,
+      "loss": 0.4431,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1986374695863747,
+      "grad_norm": 1.165526474375854,
+      "learning_rate": 9.273093684618129e-06,
+      "loss": 0.2936,
+      "step": 2041
+    },
+    {
+      "epoch": 0.19873479318734794,
+      "grad_norm": 1.6068781771010836,
+      "learning_rate": 9.272275044216981e-06,
+      "loss": 0.5125,
+      "step": 2042
+    },
+    {
+      "epoch": 0.19883211678832116,
+      "grad_norm": 1.4210491087425543,
+      "learning_rate": 9.271455979277234e-06,
+      "loss": 0.4142,
+      "step": 2043
+    },
+    {
+      "epoch": 0.19892944038929442,
+      "grad_norm": 1.6609287753373938,
+      "learning_rate": 9.270636489880283e-06,
+      "loss": 0.6728,
+      "step": 2044
+    },
+    {
+      "epoch": 0.19902676399026764,
+      "grad_norm": 1.3902108507987736,
+      "learning_rate": 9.26981657610756e-06,
+      "loss": 0.3492,
+      "step": 2045
+    },
+    {
+      "epoch": 0.19912408759124087,
+      "grad_norm": 1.6316422644879316,
+      "learning_rate": 9.268996238040537e-06,
+      "loss": 0.5029,
+      "step": 2046
+    },
+    {
+      "epoch": 0.19922141119221412,
+      "grad_norm": 1.2841836791466006,
+      "learning_rate": 9.268175475760734e-06,
+      "loss": 0.3849,
+      "step": 2047
+    },
+    {
+      "epoch": 0.19931873479318735,
+      "grad_norm": 1.319713524379575,
+      "learning_rate": 9.267354289349712e-06,
+      "loss": 0.4439,
+      "step": 2048
+    },
+    {
+      "epoch": 0.19941605839416057,
+      "grad_norm": 1.3549935774985267,
+      "learning_rate": 9.266532678889071e-06,
+      "loss": 0.4382,
+      "step": 2049
+    },
+    {
+      "epoch": 0.19951338199513383,
+      "grad_norm": 1.8518976479625036,
+      "learning_rate": 9.265710644460455e-06,
+      "loss": 0.8216,
+      "step": 2050
+    },
+    {
+      "epoch": 0.19961070559610705,
+      "grad_norm": 1.9509154982810264,
+      "learning_rate": 9.26488818614555e-06,
+      "loss": 0.4607,
+      "step": 2051
+    },
+    {
+      "epoch": 0.1997080291970803,
+      "grad_norm": 1.2954164138913125,
+      "learning_rate": 9.264065304026087e-06,
+      "loss": 0.4257,
+      "step": 2052
+    },
+    {
+      "epoch": 0.19980535279805353,
+      "grad_norm": 1.925685176039115,
+      "learning_rate": 9.26324199818383e-06,
+      "loss": 0.6025,
+      "step": 2053
+    },
+    {
+      "epoch": 0.19990267639902676,
+      "grad_norm": 1.533947029174009,
+      "learning_rate": 9.262418268700596e-06,
+      "loss": 0.5443,
+      "step": 2054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4995274594175463,
+      "learning_rate": 9.26159411565824e-06,
+      "loss": 0.5023,
+      "step": 2055
+    },
+    {
+      "epoch": 0.20009732360097324,
+      "grad_norm": 1.4350182215101954,
+      "learning_rate": 9.26076953913866e-06,
+      "loss": 0.3726,
+      "step": 2056
+    },
+    {
+      "epoch": 0.20019464720194646,
+      "grad_norm": 1.3019491914952392,
+      "learning_rate": 9.259944539223788e-06,
+      "loss": 0.4765,
+      "step": 2057
+    },
+    {
+      "epoch": 0.20029197080291972,
+      "grad_norm": 1.3884509805578256,
+      "learning_rate": 9.25911911599561e-06,
+      "loss": 0.338,
+      "step": 2058
+    },
+    {
+      "epoch": 0.20038929440389294,
+      "grad_norm": 1.488048064619486,
+      "learning_rate": 9.258293269536146e-06,
+      "loss": 0.5872,
+      "step": 2059
+    },
+    {
+      "epoch": 0.20048661800486617,
+      "grad_norm": 1.1548733119099643,
+      "learning_rate": 9.257466999927464e-06,
+      "loss": 0.3242,
+      "step": 2060
+    },
+    {
+      "epoch": 0.20058394160583942,
+      "grad_norm": 1.048222542797774,
+      "learning_rate": 9.25664030725167e-06,
+      "loss": 0.3253,
+      "step": 2061
+    },
+    {
+      "epoch": 0.20068126520681265,
+      "grad_norm": 1.211590892113714,
+      "learning_rate": 9.255813191590912e-06,
+      "loss": 0.3414,
+      "step": 2062
+    },
+    {
+      "epoch": 0.2007785888077859,
+      "grad_norm": 1.3770802107798175,
+      "learning_rate": 9.254985653027382e-06,
+      "loss": 0.4031,
+      "step": 2063
+    },
+    {
+      "epoch": 0.20087591240875913,
+      "grad_norm": 1.4503315973945832,
+      "learning_rate": 9.25415769164331e-06,
+      "loss": 0.4799,
+      "step": 2064
+    },
+    {
+      "epoch": 0.20097323600973235,
+      "grad_norm": 1.3613570222565128,
+      "learning_rate": 9.253329307520976e-06,
+      "loss": 0.3932,
+      "step": 2065
+    },
+    {
+      "epoch": 0.2010705596107056,
+      "grad_norm": 1.436956883536887,
+      "learning_rate": 9.252500500742692e-06,
+      "loss": 0.51,
+      "step": 2066
+    },
+    {
+      "epoch": 0.20116788321167883,
+      "grad_norm": 1.3042874208229347,
+      "learning_rate": 9.25167127139082e-06,
+      "loss": 0.3702,
+      "step": 2067
+    },
+    {
+      "epoch": 0.20126520681265206,
+      "grad_norm": 1.4601934649693376,
+      "learning_rate": 9.250841619547762e-06,
+      "loss": 0.3927,
+      "step": 2068
+    },
+    {
+      "epoch": 0.2013625304136253,
+      "grad_norm": 1.4877017036692342,
+      "learning_rate": 9.250011545295959e-06,
+      "loss": 0.5463,
+      "step": 2069
+    },
+    {
+      "epoch": 0.20145985401459854,
+      "grad_norm": 1.3385891837902342,
+      "learning_rate": 9.249181048717895e-06,
+      "loss": 0.3052,
+      "step": 2070
+    },
+    {
+      "epoch": 0.20155717761557176,
+      "grad_norm": 1.111892744483471,
+      "learning_rate": 9.2483501298961e-06,
+      "loss": 0.2342,
+      "step": 2071
+    },
+    {
+      "epoch": 0.20165450121654502,
+      "grad_norm": 1.4336755713622584,
+      "learning_rate": 9.247518788913141e-06,
+      "loss": 0.4416,
+      "step": 2072
+    },
+    {
+      "epoch": 0.20175182481751824,
+      "grad_norm": 1.4682039909825075,
+      "learning_rate": 9.246687025851629e-06,
+      "loss": 0.3044,
+      "step": 2073
+    },
+    {
+      "epoch": 0.2018491484184915,
+      "grad_norm": 1.1356161216510552,
+      "learning_rate": 9.245854840794217e-06,
+      "loss": 0.2913,
+      "step": 2074
+    },
+    {
+      "epoch": 0.20194647201946472,
+      "grad_norm": 1.2497989015941582,
+      "learning_rate": 9.2450222338236e-06,
+      "loss": 0.356,
+      "step": 2075
+    },
+    {
+      "epoch": 0.20204379562043795,
+      "grad_norm": 1.4662802201560914,
+      "learning_rate": 9.244189205022514e-06,
+      "loss": 0.5234,
+      "step": 2076
+    },
+    {
+      "epoch": 0.2021411192214112,
+      "grad_norm": 1.1493994388606168,
+      "learning_rate": 9.243355754473738e-06,
+      "loss": 0.3862,
+      "step": 2077
+    },
+    {
+      "epoch": 0.20223844282238443,
+      "grad_norm": 1.1352456631925198,
+      "learning_rate": 9.242521882260093e-06,
+      "loss": 0.3693,
+      "step": 2078
+    },
+    {
+      "epoch": 0.20233576642335765,
+      "grad_norm": 1.4112847797443164,
+      "learning_rate": 9.24168758846444e-06,
+      "loss": 0.4667,
+      "step": 2079
+    },
+    {
+      "epoch": 0.2024330900243309,
+      "grad_norm": 1.9587086933310962,
+      "learning_rate": 9.240852873169686e-06,
+      "loss": 0.5446,
+      "step": 2080
+    },
+    {
+      "epoch": 0.20253041362530413,
+      "grad_norm": 1.4532595336328356,
+      "learning_rate": 9.240017736458772e-06,
+      "loss": 0.56,
+      "step": 2081
+    },
+    {
+      "epoch": 0.20262773722627736,
+      "grad_norm": 1.1373158358211433,
+      "learning_rate": 9.239182178414694e-06,
+      "loss": 0.3998,
+      "step": 2082
+    },
+    {
+      "epoch": 0.2027250608272506,
+      "grad_norm": 1.4892855081953407,
+      "learning_rate": 9.238346199120473e-06,
+      "loss": 0.5564,
+      "step": 2083
+    },
+    {
+      "epoch": 0.20282238442822384,
+      "grad_norm": 1.4122351541601532,
+      "learning_rate": 9.237509798659188e-06,
+      "loss": 0.4407,
+      "step": 2084
+    },
+    {
+      "epoch": 0.2029197080291971,
+      "grad_norm": 1.266747153803517,
+      "learning_rate": 9.236672977113948e-06,
+      "loss": 0.3898,
+      "step": 2085
+    },
+    {
+      "epoch": 0.20301703163017032,
+      "grad_norm": 1.3972737248894866,
+      "learning_rate": 9.23583573456791e-06,
+      "loss": 0.4855,
+      "step": 2086
+    },
+    {
+      "epoch": 0.20311435523114355,
+      "grad_norm": 1.6424190339871019,
+      "learning_rate": 9.234998071104272e-06,
+      "loss": 0.732,
+      "step": 2087
+    },
+    {
+      "epoch": 0.2032116788321168,
+      "grad_norm": 1.4973722328869334,
+      "learning_rate": 9.234159986806275e-06,
+      "loss": 0.4796,
+      "step": 2088
+    },
+    {
+      "epoch": 0.20330900243309002,
+      "grad_norm": 1.5629802728678386,
+      "learning_rate": 9.233321481757196e-06,
+      "loss": 0.4762,
+      "step": 2089
+    },
+    {
+      "epoch": 0.20340632603406325,
+      "grad_norm": 1.5273353205689704,
+      "learning_rate": 9.23248255604036e-06,
+      "loss": 0.6446,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2035036496350365,
+      "grad_norm": 1.3835329237350877,
+      "learning_rate": 9.231643209739128e-06,
+      "loss": 0.5297,
+      "step": 2091
+    },
+    {
+      "epoch": 0.20360097323600973,
+      "grad_norm": 1.2187102873763251,
+      "learning_rate": 9.230803442936911e-06,
+      "loss": 0.3727,
+      "step": 2092
+    },
+    {
+      "epoch": 0.20369829683698296,
+      "grad_norm": 1.325749011032711,
+      "learning_rate": 9.229963255717156e-06,
+      "loss": 0.5476,
+      "step": 2093
+    },
+    {
+      "epoch": 0.2037956204379562,
+      "grad_norm": 1.1246093495598513,
+      "learning_rate": 9.229122648163351e-06,
+      "loss": 0.3309,
+      "step": 2094
+    },
+    {
+      "epoch": 0.20389294403892944,
+      "grad_norm": 1.3415111254139396,
+      "learning_rate": 9.22828162035903e-06,
+      "loss": 0.4226,
+      "step": 2095
+    },
+    {
+      "epoch": 0.2039902676399027,
+      "grad_norm": 1.2431047519820402,
+      "learning_rate": 9.227440172387766e-06,
+      "loss": 0.2364,
+      "step": 2096
+    },
+    {
+      "epoch": 0.20408759124087592,
+      "grad_norm": 1.59824202042343,
+      "learning_rate": 9.226598304333175e-06,
+      "loss": 0.5713,
+      "step": 2097
+    },
+    {
+      "epoch": 0.20418491484184914,
+      "grad_norm": 1.3718145057357327,
+      "learning_rate": 9.22575601627891e-06,
+      "loss": 0.4366,
+      "step": 2098
+    },
+    {
+      "epoch": 0.2042822384428224,
+      "grad_norm": 1.8310954422547832,
+      "learning_rate": 9.224913308308672e-06,
+      "loss": 0.4098,
+      "step": 2099
+    },
+    {
+      "epoch": 0.20437956204379562,
+      "grad_norm": 1.3433956299970118,
+      "learning_rate": 9.224070180506202e-06,
+      "loss": 0.2959,
+      "step": 2100
+    },
+    {
+      "epoch": 0.20447688564476885,
+      "grad_norm": 1.0277615122037833,
+      "learning_rate": 9.223226632955283e-06,
+      "loss": 0.265,
+      "step": 2101
+    },
+    {
+      "epoch": 0.2045742092457421,
+      "grad_norm": 1.2285380399323877,
+      "learning_rate": 9.222382665739737e-06,
+      "loss": 0.3844,
+      "step": 2102
+    },
+    {
+      "epoch": 0.20467153284671533,
+      "grad_norm": 1.1151094116106592,
+      "learning_rate": 9.221538278943432e-06,
+      "loss": 0.2461,
+      "step": 2103
+    },
+    {
+      "epoch": 0.20476885644768855,
+      "grad_norm": 1.5239102143699876,
+      "learning_rate": 9.22069347265027e-06,
+      "loss": 0.4239,
+      "step": 2104
+    },
+    {
+      "epoch": 0.2048661800486618,
+      "grad_norm": 1.6502658051911525,
+      "learning_rate": 9.219848246944206e-06,
+      "loss": 0.6723,
+      "step": 2105
+    },
+    {
+      "epoch": 0.20496350364963503,
+      "grad_norm": 1.638974040274465,
+      "learning_rate": 9.219002601909229e-06,
+      "loss": 0.5068,
+      "step": 2106
+    },
+    {
+      "epoch": 0.20506082725060829,
+      "grad_norm": 1.4649352184984061,
+      "learning_rate": 9.218156537629368e-06,
+      "loss": 0.4698,
+      "step": 2107
+    },
+    {
+      "epoch": 0.2051581508515815,
+      "grad_norm": 1.5070786345583258,
+      "learning_rate": 9.217310054188699e-06,
+      "loss": 0.4654,
+      "step": 2108
+    },
+    {
+      "epoch": 0.20525547445255474,
+      "grad_norm": 1.2480947756940115,
+      "learning_rate": 9.216463151671338e-06,
+      "loss": 0.3614,
+      "step": 2109
+    },
+    {
+      "epoch": 0.205352798053528,
+      "grad_norm": 1.6536121595263205,
+      "learning_rate": 9.215615830161443e-06,
+      "loss": 0.5872,
+      "step": 2110
+    },
+    {
+      "epoch": 0.20545012165450122,
+      "grad_norm": 1.5559546132859907,
+      "learning_rate": 9.214768089743211e-06,
+      "loss": 0.5098,
+      "step": 2111
+    },
+    {
+      "epoch": 0.20554744525547444,
+      "grad_norm": 1.5691593927804695,
+      "learning_rate": 9.213919930500884e-06,
+      "loss": 0.3845,
+      "step": 2112
+    },
+    {
+      "epoch": 0.2056447688564477,
+      "grad_norm": 1.4385010923740136,
+      "learning_rate": 9.213071352518744e-06,
+      "loss": 0.4035,
+      "step": 2113
+    },
+    {
+      "epoch": 0.20574209245742092,
+      "grad_norm": 1.2415148755341134,
+      "learning_rate": 9.212222355881111e-06,
+      "loss": 0.2503,
+      "step": 2114
+    },
+    {
+      "epoch": 0.20583941605839415,
+      "grad_norm": 1.597224767194554,
+      "learning_rate": 9.211372940672356e-06,
+      "loss": 0.3831,
+      "step": 2115
+    },
+    {
+      "epoch": 0.2059367396593674,
+      "grad_norm": 1.3936071245663937,
+      "learning_rate": 9.210523106976884e-06,
+      "loss": 0.3664,
+      "step": 2116
+    },
+    {
+      "epoch": 0.20603406326034063,
+      "grad_norm": 1.4335641468120297,
+      "learning_rate": 9.209672854879142e-06,
+      "loss": 0.3182,
+      "step": 2117
+    },
+    {
+      "epoch": 0.20613138686131388,
+      "grad_norm": 1.2544256067640176,
+      "learning_rate": 9.20882218446362e-06,
+      "loss": 0.2678,
+      "step": 2118
+    },
+    {
+      "epoch": 0.2062287104622871,
+      "grad_norm": 1.4867246001264303,
+      "learning_rate": 9.207971095814852e-06,
+      "loss": 0.4934,
+      "step": 2119
+    },
+    {
+      "epoch": 0.20632603406326033,
+      "grad_norm": 1.5387304887069146,
+      "learning_rate": 9.207119589017408e-06,
+      "loss": 0.4552,
+      "step": 2120
+    },
+    {
+      "epoch": 0.2064233576642336,
+      "grad_norm": 1.507156387441411,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 0.4209,
+      "step": 2121
+    },
+    {
+      "epoch": 0.2065206812652068,
+      "grad_norm": 1.3407732350308024,
+      "learning_rate": 9.205415321315e-06,
+      "loss": 0.4256,
+      "step": 2122
+    },
+    {
+      "epoch": 0.20661800486618004,
+      "grad_norm": 1.6313949345186305,
+      "learning_rate": 9.20456256057939e-06,
+      "loss": 0.4727,
+      "step": 2123
+    },
+    {
+      "epoch": 0.2067153284671533,
+      "grad_norm": 1.695026004332969,
+      "learning_rate": 9.203709382033814e-06,
+      "loss": 0.6547,
+      "step": 2124
+    },
+    {
+      "epoch": 0.20681265206812652,
+      "grad_norm": 1.5677721722384952,
+      "learning_rate": 9.202855785763053e-06,
+      "loss": 0.4469,
+      "step": 2125
+    },
+    {
+      "epoch": 0.20690997566909974,
+      "grad_norm": 1.4276579746412523,
+      "learning_rate": 9.202001771851928e-06,
+      "loss": 0.4511,
+      "step": 2126
+    },
+    {
+      "epoch": 0.207007299270073,
+      "grad_norm": 1.365652083209099,
+      "learning_rate": 9.201147340385304e-06,
+      "loss": 0.4435,
+      "step": 2127
+    },
+    {
+      "epoch": 0.20710462287104622,
+      "grad_norm": 1.4014399599326692,
+      "learning_rate": 9.200292491448086e-06,
+      "loss": 0.4017,
+      "step": 2128
+    },
+    {
+      "epoch": 0.20720194647201948,
+      "grad_norm": 1.4131798281318602,
+      "learning_rate": 9.199437225125223e-06,
+      "loss": 0.2781,
+      "step": 2129
+    },
+    {
+      "epoch": 0.2072992700729927,
+      "grad_norm": 1.3392698432345278,
+      "learning_rate": 9.198581541501702e-06,
+      "loss": 0.3576,
+      "step": 2130
+    },
+    {
+      "epoch": 0.20739659367396593,
+      "grad_norm": 1.2859171090531423,
+      "learning_rate": 9.197725440662552e-06,
+      "loss": 0.4505,
+      "step": 2131
+    },
+    {
+      "epoch": 0.20749391727493918,
+      "grad_norm": 1.3075221898254676,
+      "learning_rate": 9.196868922692845e-06,
+      "loss": 0.42,
+      "step": 2132
+    },
+    {
+      "epoch": 0.2075912408759124,
+      "grad_norm": 1.3120969425940014,
+      "learning_rate": 9.196011987677693e-06,
+      "loss": 0.3918,
+      "step": 2133
+    },
+    {
+      "epoch": 0.20768856447688563,
+      "grad_norm": 1.2917866907447901,
+      "learning_rate": 9.19515463570225e-06,
+      "loss": 0.4515,
+      "step": 2134
+    },
+    {
+      "epoch": 0.2077858880778589,
+      "grad_norm": 1.4964227937052923,
+      "learning_rate": 9.194296866851714e-06,
+      "loss": 0.4007,
+      "step": 2135
+    },
+    {
+      "epoch": 0.2078832116788321,
+      "grad_norm": 1.4096694486456338,
+      "learning_rate": 9.19343868121132e-06,
+      "loss": 0.5684,
+      "step": 2136
+    },
+    {
+      "epoch": 0.20798053527980534,
+      "grad_norm": 1.1303877036272907,
+      "learning_rate": 9.192580078866346e-06,
+      "loss": 0.2661,
+      "step": 2137
+    },
+    {
+      "epoch": 0.2080778588807786,
+      "grad_norm": 1.4056619474271335,
+      "learning_rate": 9.191721059902112e-06,
+      "loss": 0.4174,
+      "step": 2138
+    },
+    {
+      "epoch": 0.20817518248175182,
+      "grad_norm": 1.7142064467904727,
+      "learning_rate": 9.190861624403981e-06,
+      "loss": 0.4453,
+      "step": 2139
+    },
+    {
+      "epoch": 0.20827250608272507,
+      "grad_norm": 1.3293557691236777,
+      "learning_rate": 9.190001772457356e-06,
+      "loss": 0.4541,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2083698296836983,
+      "grad_norm": 1.6131133576379075,
+      "learning_rate": 9.189141504147676e-06,
+      "loss": 0.3751,
+      "step": 2141
+    },
+    {
+      "epoch": 0.20846715328467152,
+      "grad_norm": 1.509737357483189,
+      "learning_rate": 9.188280819560431e-06,
+      "loss": 0.4757,
+      "step": 2142
+    },
+    {
+      "epoch": 0.20856447688564478,
+      "grad_norm": 1.479538114231473,
+      "learning_rate": 9.187419718781149e-06,
+      "loss": 0.3243,
+      "step": 2143
+    },
+    {
+      "epoch": 0.208661800486618,
+      "grad_norm": 1.4973982658919327,
+      "learning_rate": 9.186558201895395e-06,
+      "loss": 0.3732,
+      "step": 2144
+    },
+    {
+      "epoch": 0.20875912408759123,
+      "grad_norm": 1.5121453838943797,
+      "learning_rate": 9.185696268988777e-06,
+      "loss": 0.5435,
+      "step": 2145
+    },
+    {
+      "epoch": 0.20885644768856448,
+      "grad_norm": 1.7349033410138828,
+      "learning_rate": 9.18483392014695e-06,
+      "loss": 0.6415,
+      "step": 2146
+    },
+    {
+      "epoch": 0.2089537712895377,
+      "grad_norm": 1.4812330220855032,
+      "learning_rate": 9.183971155455602e-06,
+      "loss": 0.4961,
+      "step": 2147
+    },
+    {
+      "epoch": 0.20905109489051094,
+      "grad_norm": 1.5121767597167877,
+      "learning_rate": 9.183107975000472e-06,
+      "loss": 0.5298,
+      "step": 2148
+    },
+    {
+      "epoch": 0.2091484184914842,
+      "grad_norm": 1.5424817825799644,
+      "learning_rate": 9.18224437886733e-06,
+      "loss": 0.4577,
+      "step": 2149
+    },
+    {
+      "epoch": 0.20924574209245742,
+      "grad_norm": 1.2733853569354763,
+      "learning_rate": 9.181380367141991e-06,
+      "loss": 0.3306,
+      "step": 2150
+    },
+    {
+      "epoch": 0.20934306569343067,
+      "grad_norm": 1.1384650904715041,
+      "learning_rate": 9.180515939910317e-06,
+      "loss": 0.3831,
+      "step": 2151
+    },
+    {
+      "epoch": 0.2094403892944039,
+      "grad_norm": 1.3798308474076018,
+      "learning_rate": 9.179651097258204e-06,
+      "loss": 0.4629,
+      "step": 2152
+    },
+    {
+      "epoch": 0.20953771289537712,
+      "grad_norm": 1.4059733648531154,
+      "learning_rate": 9.178785839271593e-06,
+      "loss": 0.4526,
+      "step": 2153
+    },
+    {
+      "epoch": 0.20963503649635037,
+      "grad_norm": 1.581039004516103,
+      "learning_rate": 9.177920166036464e-06,
+      "loss": 0.5397,
+      "step": 2154
+    },
+    {
+      "epoch": 0.2097323600973236,
+      "grad_norm": 1.4851118969101265,
+      "learning_rate": 9.17705407763884e-06,
+      "loss": 0.5052,
+      "step": 2155
+    },
+    {
+      "epoch": 0.20982968369829683,
+      "grad_norm": 1.3633687775503893,
+      "learning_rate": 9.176187574164785e-06,
+      "loss": 0.4427,
+      "step": 2156
+    },
+    {
+      "epoch": 0.20992700729927008,
+      "grad_norm": 1.360319094739405,
+      "learning_rate": 9.175320655700407e-06,
+      "loss": 0.3649,
+      "step": 2157
+    },
+    {
+      "epoch": 0.2100243309002433,
+      "grad_norm": 1.3829673206277566,
+      "learning_rate": 9.174453322331844e-06,
+      "loss": 0.3536,
+      "step": 2158
+    },
+    {
+      "epoch": 0.21012165450121653,
+      "grad_norm": 1.5804059757696094,
+      "learning_rate": 9.173585574145292e-06,
+      "loss": 0.5937,
+      "step": 2159
+    },
+    {
+      "epoch": 0.21021897810218979,
+      "grad_norm": 1.4991084469228289,
+      "learning_rate": 9.172717411226975e-06,
+      "loss": 0.3523,
+      "step": 2160
+    },
+    {
+      "epoch": 0.210316301703163,
+      "grad_norm": 1.4762289487935065,
+      "learning_rate": 9.171848833663165e-06,
+      "loss": 0.4991,
+      "step": 2161
+    },
+    {
+      "epoch": 0.21041362530413626,
+      "grad_norm": 1.4858484283610454,
+      "learning_rate": 9.17097984154017e-06,
+      "loss": 0.5153,
+      "step": 2162
+    },
+    {
+      "epoch": 0.2105109489051095,
+      "grad_norm": 1.2647097068290445,
+      "learning_rate": 9.170110434944345e-06,
+      "loss": 0.3193,
+      "step": 2163
+    },
+    {
+      "epoch": 0.21060827250608272,
+      "grad_norm": 1.6889738075479466,
+      "learning_rate": 9.169240613962086e-06,
+      "loss": 0.4755,
+      "step": 2164
+    },
+    {
+      "epoch": 0.21070559610705597,
+      "grad_norm": 1.6464662019172414,
+      "learning_rate": 9.168370378679821e-06,
+      "loss": 0.5303,
+      "step": 2165
+    },
+    {
+      "epoch": 0.2108029197080292,
+      "grad_norm": 1.287927301108519,
+      "learning_rate": 9.16749972918403e-06,
+      "loss": 0.3231,
+      "step": 2166
+    },
+    {
+      "epoch": 0.21090024330900242,
+      "grad_norm": 1.378935902738664,
+      "learning_rate": 9.16662866556123e-06,
+      "loss": 0.4654,
+      "step": 2167
+    },
+    {
+      "epoch": 0.21099756690997568,
+      "grad_norm": 1.415652566603492,
+      "learning_rate": 9.16575718789798e-06,
+      "loss": 0.42,
+      "step": 2168
+    },
+    {
+      "epoch": 0.2110948905109489,
+      "grad_norm": 1.189498123796033,
+      "learning_rate": 9.164885296280875e-06,
+      "loss": 0.3529,
+      "step": 2169
+    },
+    {
+      "epoch": 0.21119221411192213,
+      "grad_norm": 1.5371351227791108,
+      "learning_rate": 9.16401299079656e-06,
+      "loss": 0.4679,
+      "step": 2170
+    },
+    {
+      "epoch": 0.21128953771289538,
+      "grad_norm": 1.2493790037654902,
+      "learning_rate": 9.163140271531714e-06,
+      "loss": 0.3793,
+      "step": 2171
+    },
+    {
+      "epoch": 0.2113868613138686,
+      "grad_norm": 1.3836947713855836,
+      "learning_rate": 9.16226713857306e-06,
+      "loss": 0.436,
+      "step": 2172
+    },
+    {
+      "epoch": 0.21148418491484186,
+      "grad_norm": 1.583280621035993,
+      "learning_rate": 9.161393592007364e-06,
+      "loss": 0.5673,
+      "step": 2173
+    },
+    {
+      "epoch": 0.2115815085158151,
+      "grad_norm": 1.336076606512916,
+      "learning_rate": 9.160519631921427e-06,
+      "loss": 0.418,
+      "step": 2174
+    },
+    {
+      "epoch": 0.2116788321167883,
+      "grad_norm": 1.5539773056945747,
+      "learning_rate": 9.159645258402098e-06,
+      "loss": 0.4417,
+      "step": 2175
+    },
+    {
+      "epoch": 0.21177615571776157,
+      "grad_norm": 1.35099904216899,
+      "learning_rate": 9.158770471536261e-06,
+      "loss": 0.4389,
+      "step": 2176
+    },
+    {
+      "epoch": 0.2118734793187348,
+      "grad_norm": 1.5960801985245197,
+      "learning_rate": 9.157895271410848e-06,
+      "loss": 0.4444,
+      "step": 2177
+    },
+    {
+      "epoch": 0.21197080291970802,
+      "grad_norm": 1.343338393224711,
+      "learning_rate": 9.157019658112825e-06,
+      "loss": 0.3867,
+      "step": 2178
+    },
+    {
+      "epoch": 0.21206812652068127,
+      "grad_norm": 1.573040163695098,
+      "learning_rate": 9.156143631729205e-06,
+      "loss": 0.5564,
+      "step": 2179
+    },
+    {
+      "epoch": 0.2121654501216545,
+      "grad_norm": 1.477194998770335,
+      "learning_rate": 9.155267192347037e-06,
+      "loss": 0.5053,
+      "step": 2180
+    },
+    {
+      "epoch": 0.21226277372262772,
+      "grad_norm": 1.4697445687746653,
+      "learning_rate": 9.154390340053414e-06,
+      "loss": 0.4462,
+      "step": 2181
+    },
+    {
+      "epoch": 0.21236009732360098,
+      "grad_norm": 1.2383233673923462,
+      "learning_rate": 9.15351307493547e-06,
+      "loss": 0.4023,
+      "step": 2182
+    },
+    {
+      "epoch": 0.2124574209245742,
+      "grad_norm": 1.73929160255024,
+      "learning_rate": 9.152635397080377e-06,
+      "loss": 0.456,
+      "step": 2183
+    },
+    {
+      "epoch": 0.21255474452554746,
+      "grad_norm": 1.814215933055299,
+      "learning_rate": 9.151757306575354e-06,
+      "loss": 0.5283,
+      "step": 2184
+    },
+    {
+      "epoch": 0.21265206812652068,
+      "grad_norm": 1.440140413882406,
+      "learning_rate": 9.150878803507655e-06,
+      "loss": 0.4754,
+      "step": 2185
+    },
+    {
+      "epoch": 0.2127493917274939,
+      "grad_norm": 1.4991761170210094,
+      "learning_rate": 9.149999887964577e-06,
+      "loss": 0.4244,
+      "step": 2186
+    },
+    {
+      "epoch": 0.21284671532846716,
+      "grad_norm": 1.6045542244692401,
+      "learning_rate": 9.149120560033461e-06,
+      "loss": 0.4149,
+      "step": 2187
+    },
+    {
+      "epoch": 0.2129440389294404,
+      "grad_norm": 1.6999406355422166,
+      "learning_rate": 9.148240819801684e-06,
+      "loss": 0.7227,
+      "step": 2188
+    },
+    {
+      "epoch": 0.2130413625304136,
+      "grad_norm": 1.5383336234101048,
+      "learning_rate": 9.147360667356667e-06,
+      "loss": 0.4102,
+      "step": 2189
+    },
+    {
+      "epoch": 0.21313868613138687,
+      "grad_norm": 1.3100772476716567,
+      "learning_rate": 9.146480102785871e-06,
+      "loss": 0.4001,
+      "step": 2190
+    },
+    {
+      "epoch": 0.2132360097323601,
+      "grad_norm": 1.2113504505529646,
+      "learning_rate": 9.1455991261768e-06,
+      "loss": 0.3906,
+      "step": 2191
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 2.1524156395732996,
+      "learning_rate": 9.144717737616994e-06,
+      "loss": 0.3722,
+      "step": 2192
+    },
+    {
+      "epoch": 0.21343065693430657,
+      "grad_norm": 1.3156410053212892,
+      "learning_rate": 9.143835937194039e-06,
+      "loss": 0.414,
+      "step": 2193
+    },
+    {
+      "epoch": 0.2135279805352798,
+      "grad_norm": 1.382537469614808,
+      "learning_rate": 9.14295372499556e-06,
+      "loss": 0.3687,
+      "step": 2194
+    },
+    {
+      "epoch": 0.21362530413625305,
+      "grad_norm": 1.4106617705657403,
+      "learning_rate": 9.142071101109224e-06,
+      "loss": 0.2515,
+      "step": 2195
+    },
+    {
+      "epoch": 0.21372262773722628,
+      "grad_norm": 1.4292530170893925,
+      "learning_rate": 9.141188065622736e-06,
+      "loss": 0.4671,
+      "step": 2196
+    },
+    {
+      "epoch": 0.2138199513381995,
+      "grad_norm": 1.371262803483025,
+      "learning_rate": 9.140304618623844e-06,
+      "loss": 0.4397,
+      "step": 2197
+    },
+    {
+      "epoch": 0.21391727493917276,
+      "grad_norm": 1.3337172412513854,
+      "learning_rate": 9.13942076020034e-06,
+      "loss": 0.4518,
+      "step": 2198
+    },
+    {
+      "epoch": 0.21401459854014598,
+      "grad_norm": 1.195478639712577,
+      "learning_rate": 9.138536490440046e-06,
+      "loss": 0.3236,
+      "step": 2199
+    },
+    {
+      "epoch": 0.2141119221411192,
+      "grad_norm": 1.6207375008593756,
+      "learning_rate": 9.13765180943084e-06,
+      "loss": 0.5147,
+      "step": 2200
+    },
+    {
+      "epoch": 0.21420924574209246,
+      "grad_norm": 1.457360033672521,
+      "learning_rate": 9.136766717260631e-06,
+      "loss": 0.3228,
+      "step": 2201
+    },
+    {
+      "epoch": 0.2143065693430657,
+      "grad_norm": 1.2314544120773039,
+      "learning_rate": 9.13588121401737e-06,
+      "loss": 0.3413,
+      "step": 2202
+    },
+    {
+      "epoch": 0.21440389294403894,
+      "grad_norm": 1.3614880154600904,
+      "learning_rate": 9.13499529978905e-06,
+      "loss": 0.3902,
+      "step": 2203
+    },
+    {
+      "epoch": 0.21450121654501217,
+      "grad_norm": 1.3431981306372034,
+      "learning_rate": 9.134108974663707e-06,
+      "loss": 0.4893,
+      "step": 2204
+    },
+    {
+      "epoch": 0.2145985401459854,
+      "grad_norm": 1.346114362934121,
+      "learning_rate": 9.133222238729414e-06,
+      "loss": 0.4195,
+      "step": 2205
+    },
+    {
+      "epoch": 0.21469586374695865,
+      "grad_norm": 1.2405202461045035,
+      "learning_rate": 9.132335092074285e-06,
+      "loss": 0.4373,
+      "step": 2206
+    },
+    {
+      "epoch": 0.21479318734793187,
+      "grad_norm": 1.2952176269832685,
+      "learning_rate": 9.131447534786478e-06,
+      "loss": 0.3253,
+      "step": 2207
+    },
+    {
+      "epoch": 0.2148905109489051,
+      "grad_norm": 1.3497804127584312,
+      "learning_rate": 9.130559566954191e-06,
+      "loss": 0.4401,
+      "step": 2208
+    },
+    {
+      "epoch": 0.21498783454987835,
+      "grad_norm": 1.6094605506454212,
+      "learning_rate": 9.129671188665661e-06,
+      "loss": 0.5943,
+      "step": 2209
+    },
+    {
+      "epoch": 0.21508515815085158,
+      "grad_norm": 1.7393737788578179,
+      "learning_rate": 9.128782400009167e-06,
+      "loss": 0.6832,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2151824817518248,
+      "grad_norm": 1.2888456219960003,
+      "learning_rate": 9.127893201073028e-06,
+      "loss": 0.4449,
+      "step": 2211
+    },
+    {
+      "epoch": 0.21527980535279806,
+      "grad_norm": 1.6231451452368957,
+      "learning_rate": 9.127003591945605e-06,
+      "loss": 0.6579,
+      "step": 2212
+    },
+    {
+      "epoch": 0.21537712895377129,
+      "grad_norm": 1.4013330754504585,
+      "learning_rate": 9.126113572715296e-06,
+      "loss": 0.5072,
+      "step": 2213
+    },
+    {
+      "epoch": 0.21547445255474454,
+      "grad_norm": 1.1928349667862592,
+      "learning_rate": 9.125223143470547e-06,
+      "loss": 0.2896,
+      "step": 2214
+    },
+    {
+      "epoch": 0.21557177615571776,
+      "grad_norm": 1.3027255903002162,
+      "learning_rate": 9.124332304299838e-06,
+      "loss": 0.3076,
+      "step": 2215
+    },
+    {
+      "epoch": 0.215669099756691,
+      "grad_norm": 1.6527022746103417,
+      "learning_rate": 9.123441055291694e-06,
+      "loss": 0.4688,
+      "step": 2216
+    },
+    {
+      "epoch": 0.21576642335766424,
+      "grad_norm": 1.3197927862863625,
+      "learning_rate": 9.122549396534676e-06,
+      "loss": 0.318,
+      "step": 2217
+    },
+    {
+      "epoch": 0.21586374695863747,
+      "grad_norm": 1.5297610770776902,
+      "learning_rate": 9.121657328117392e-06,
+      "loss": 0.6176,
+      "step": 2218
+    },
+    {
+      "epoch": 0.2159610705596107,
+      "grad_norm": 1.338041823259507,
+      "learning_rate": 9.120764850128486e-06,
+      "loss": 0.3941,
+      "step": 2219
+    },
+    {
+      "epoch": 0.21605839416058395,
+      "grad_norm": 1.200858421054794,
+      "learning_rate": 9.119871962656644e-06,
+      "loss": 0.3758,
+      "step": 2220
+    },
+    {
+      "epoch": 0.21615571776155718,
+      "grad_norm": 1.5023816592412242,
+      "learning_rate": 9.118978665790592e-06,
+      "loss": 0.5032,
+      "step": 2221
+    },
+    {
+      "epoch": 0.2162530413625304,
+      "grad_norm": 1.2258656459952086,
+      "learning_rate": 9.118084959619099e-06,
+      "loss": 0.4489,
+      "step": 2222
+    },
+    {
+      "epoch": 0.21635036496350366,
+      "grad_norm": 1.717063075964899,
+      "learning_rate": 9.117190844230971e-06,
+      "loss": 0.7762,
+      "step": 2223
+    },
+    {
+      "epoch": 0.21644768856447688,
+      "grad_norm": 1.210140433555958,
+      "learning_rate": 9.11629631971506e-06,
+      "loss": 0.4431,
+      "step": 2224
+    },
+    {
+      "epoch": 0.21654501216545013,
+      "grad_norm": 1.4188251693910732,
+      "learning_rate": 9.115401386160252e-06,
+      "loss": 0.3495,
+      "step": 2225
+    },
+    {
+      "epoch": 0.21664233576642336,
+      "grad_norm": 2.073136961715272,
+      "learning_rate": 9.11450604365548e-06,
+      "loss": 0.4268,
+      "step": 2226
+    },
+    {
+      "epoch": 0.2167396593673966,
+      "grad_norm": 1.5265588328884594,
+      "learning_rate": 9.113610292289714e-06,
+      "loss": 0.4303,
+      "step": 2227
+    },
+    {
+      "epoch": 0.21683698296836984,
+      "grad_norm": 1.3220401272995868,
+      "learning_rate": 9.112714132151963e-06,
+      "loss": 0.4221,
+      "step": 2228
+    },
+    {
+      "epoch": 0.21693430656934307,
+      "grad_norm": 1.4088441022230214,
+      "learning_rate": 9.111817563331282e-06,
+      "loss": 0.1886,
+      "step": 2229
+    },
+    {
+      "epoch": 0.2170316301703163,
+      "grad_norm": 1.3947572498286958,
+      "learning_rate": 9.110920585916763e-06,
+      "loss": 0.353,
+      "step": 2230
+    },
+    {
+      "epoch": 0.21712895377128955,
+      "grad_norm": 1.2369368803593181,
+      "learning_rate": 9.110023199997537e-06,
+      "loss": 0.2576,
+      "step": 2231
+    },
+    {
+      "epoch": 0.21722627737226277,
+      "grad_norm": 1.1860471672244592,
+      "learning_rate": 9.10912540566278e-06,
+      "loss": 0.3994,
+      "step": 2232
+    },
+    {
+      "epoch": 0.217323600973236,
+      "grad_norm": 1.309576411449957,
+      "learning_rate": 9.108227203001708e-06,
+      "loss": 0.4453,
+      "step": 2233
+    },
+    {
+      "epoch": 0.21742092457420925,
+      "grad_norm": 1.6554896930775824,
+      "learning_rate": 9.10732859210357e-06,
+      "loss": 0.589,
+      "step": 2234
+    },
+    {
+      "epoch": 0.21751824817518248,
+      "grad_norm": 1.761859219992272,
+      "learning_rate": 9.106429573057666e-06,
+      "loss": 0.726,
+      "step": 2235
+    },
+    {
+      "epoch": 0.21761557177615573,
+      "grad_norm": 1.35833156484165,
+      "learning_rate": 9.105530145953335e-06,
+      "loss": 0.4012,
+      "step": 2236
+    },
+    {
+      "epoch": 0.21771289537712896,
+      "grad_norm": 3.4502529438559884,
+      "learning_rate": 9.104630310879944e-06,
+      "loss": 0.4621,
+      "step": 2237
+    },
+    {
+      "epoch": 0.21781021897810218,
+      "grad_norm": 1.3357957463599541,
+      "learning_rate": 9.103730067926922e-06,
+      "loss": 0.317,
+      "step": 2238
+    },
+    {
+      "epoch": 0.21790754257907544,
+      "grad_norm": 1.3566642568052916,
+      "learning_rate": 9.102829417183716e-06,
+      "loss": 0.4245,
+      "step": 2239
+    },
+    {
+      "epoch": 0.21800486618004866,
+      "grad_norm": 1.673808040965782,
+      "learning_rate": 9.10192835873983e-06,
+      "loss": 0.6908,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2181021897810219,
+      "grad_norm": 1.8194308130790637,
+      "learning_rate": 9.101026892684804e-06,
+      "loss": 0.5157,
+      "step": 2241
+    },
+    {
+      "epoch": 0.21819951338199514,
+      "grad_norm": 1.4443029228393756,
+      "learning_rate": 9.100125019108214e-06,
+      "loss": 0.5417,
+      "step": 2242
+    },
+    {
+      "epoch": 0.21829683698296837,
+      "grad_norm": 1.4594341846039764,
+      "learning_rate": 9.099222738099682e-06,
+      "loss": 0.4297,
+      "step": 2243
+    },
+    {
+      "epoch": 0.2183941605839416,
+      "grad_norm": 1.3121064822320374,
+      "learning_rate": 9.098320049748864e-06,
+      "loss": 0.4646,
+      "step": 2244
+    },
+    {
+      "epoch": 0.21849148418491485,
+      "grad_norm": 1.5596348242175504,
+      "learning_rate": 9.097416954145467e-06,
+      "loss": 0.4877,
+      "step": 2245
+    },
+    {
+      "epoch": 0.21858880778588807,
+      "grad_norm": 1.1835003302943965,
+      "learning_rate": 9.096513451379225e-06,
+      "loss": 0.3548,
+      "step": 2246
+    },
+    {
+      "epoch": 0.21868613138686133,
+      "grad_norm": 1.4956699498169375,
+      "learning_rate": 9.095609541539925e-06,
+      "loss": 0.3958,
+      "step": 2247
+    },
+    {
+      "epoch": 0.21878345498783455,
+      "grad_norm": 1.3761247023142853,
+      "learning_rate": 9.094705224717388e-06,
+      "loss": 0.4076,
+      "step": 2248
+    },
+    {
+      "epoch": 0.21888077858880778,
+      "grad_norm": 1.2940624946938768,
+      "learning_rate": 9.093800501001476e-06,
+      "loss": 0.4989,
+      "step": 2249
+    },
+    {
+      "epoch": 0.21897810218978103,
+      "grad_norm": 1.1389229499303237,
+      "learning_rate": 9.092895370482091e-06,
+      "loss": 0.332,
+      "step": 2250
+    },
+    {
+      "epoch": 0.21907542579075426,
+      "grad_norm": 1.5338979130860617,
+      "learning_rate": 9.091989833249179e-06,
+      "loss": 0.5609,
+      "step": 2251
+    },
+    {
+      "epoch": 0.21917274939172748,
+      "grad_norm": 1.3736786128370664,
+      "learning_rate": 9.091083889392721e-06,
+      "loss": 0.3767,
+      "step": 2252
+    },
+    {
+      "epoch": 0.21927007299270074,
+      "grad_norm": 1.6001218689759074,
+      "learning_rate": 9.090177539002743e-06,
+      "loss": 0.5709,
+      "step": 2253
+    },
+    {
+      "epoch": 0.21936739659367396,
+      "grad_norm": 1.2578364778685514,
+      "learning_rate": 9.089270782169308e-06,
+      "loss": 0.3796,
+      "step": 2254
+    },
+    {
+      "epoch": 0.2194647201946472,
+      "grad_norm": 1.5508865589735865,
+      "learning_rate": 9.088363618982523e-06,
+      "loss": 0.5947,
+      "step": 2255
+    },
+    {
+      "epoch": 0.21956204379562044,
+      "grad_norm": 1.2646857650137902,
+      "learning_rate": 9.08745604953253e-06,
+      "loss": 0.3024,
+      "step": 2256
+    },
+    {
+      "epoch": 0.21965936739659367,
+      "grad_norm": 1.1168071392771144,
+      "learning_rate": 9.08654807390952e-06,
+      "loss": 0.3113,
+      "step": 2257
+    },
+    {
+      "epoch": 0.21975669099756692,
+      "grad_norm": 1.238369237619726,
+      "learning_rate": 9.085639692203713e-06,
+      "loss": 0.2179,
+      "step": 2258
+    },
+    {
+      "epoch": 0.21985401459854015,
+      "grad_norm": 1.2485790759653945,
+      "learning_rate": 9.084730904505381e-06,
+      "loss": 0.3763,
+      "step": 2259
+    },
+    {
+      "epoch": 0.21995133819951337,
+      "grad_norm": 1.6082877032407055,
+      "learning_rate": 9.083821710904827e-06,
+      "loss": 0.3831,
+      "step": 2260
+    },
+    {
+      "epoch": 0.22004866180048663,
+      "grad_norm": 1.3213256018887491,
+      "learning_rate": 9.082912111492401e-06,
+      "loss": 0.4091,
+      "step": 2261
+    },
+    {
+      "epoch": 0.22014598540145985,
+      "grad_norm": 1.5899440724355371,
+      "learning_rate": 9.08200210635849e-06,
+      "loss": 0.4491,
+      "step": 2262
+    },
+    {
+      "epoch": 0.22024330900243308,
+      "grad_norm": 1.30089465497526,
+      "learning_rate": 9.081091695593518e-06,
+      "loss": 0.3762,
+      "step": 2263
+    },
+    {
+      "epoch": 0.22034063260340633,
+      "grad_norm": 1.5403984971127525,
+      "learning_rate": 9.080180879287957e-06,
+      "loss": 0.438,
+      "step": 2264
+    },
+    {
+      "epoch": 0.22043795620437956,
+      "grad_norm": 1.5500984898931875,
+      "learning_rate": 9.079269657532312e-06,
+      "loss": 0.398,
+      "step": 2265
+    },
+    {
+      "epoch": 0.22053527980535279,
+      "grad_norm": 1.4834461719298844,
+      "learning_rate": 9.078358030417136e-06,
+      "loss": 0.6175,
+      "step": 2266
+    },
+    {
+      "epoch": 0.22063260340632604,
+      "grad_norm": 1.3553003212010182,
+      "learning_rate": 9.077445998033015e-06,
+      "loss": 0.2719,
+      "step": 2267
+    },
+    {
+      "epoch": 0.22072992700729926,
+      "grad_norm": 1.573783871238475,
+      "learning_rate": 9.07653356047058e-06,
+      "loss": 0.2328,
+      "step": 2268
+    },
+    {
+      "epoch": 0.22082725060827252,
+      "grad_norm": 1.54928316645126,
+      "learning_rate": 9.075620717820498e-06,
+      "loss": 0.3514,
+      "step": 2269
+    },
+    {
+      "epoch": 0.22092457420924574,
+      "grad_norm": 1.3616253433976528,
+      "learning_rate": 9.07470747017348e-06,
+      "loss": 0.4636,
+      "step": 2270
+    },
+    {
+      "epoch": 0.22102189781021897,
+      "grad_norm": 1.6741713680481711,
+      "learning_rate": 9.073793817620277e-06,
+      "loss": 0.6321,
+      "step": 2271
+    },
+    {
+      "epoch": 0.22111922141119222,
+      "grad_norm": 1.3794305685281492,
+      "learning_rate": 9.07287976025168e-06,
+      "loss": 0.3172,
+      "step": 2272
+    },
+    {
+      "epoch": 0.22121654501216545,
+      "grad_norm": 1.362894347632133,
+      "learning_rate": 9.071965298158516e-06,
+      "loss": 0.3989,
+      "step": 2273
+    },
+    {
+      "epoch": 0.22131386861313868,
+      "grad_norm": 1.4233131262232992,
+      "learning_rate": 9.071050431431658e-06,
+      "loss": 0.4922,
+      "step": 2274
+    },
+    {
+      "epoch": 0.22141119221411193,
+      "grad_norm": 1.4905332812995968,
+      "learning_rate": 9.070135160162016e-06,
+      "loss": 0.3952,
+      "step": 2275
+    },
+    {
+      "epoch": 0.22150851581508516,
+      "grad_norm": 1.4389307945528345,
+      "learning_rate": 9.069219484440541e-06,
+      "loss": 0.4364,
+      "step": 2276
+    },
+    {
+      "epoch": 0.22160583941605838,
+      "grad_norm": 1.4796907096594347,
+      "learning_rate": 9.068303404358226e-06,
+      "loss": 0.4842,
+      "step": 2277
+    },
+    {
+      "epoch": 0.22170316301703163,
+      "grad_norm": 1.6561415294899449,
+      "learning_rate": 9.0673869200061e-06,
+      "loss": 0.5595,
+      "step": 2278
+    },
+    {
+      "epoch": 0.22180048661800486,
+      "grad_norm": 1.4198474890784685,
+      "learning_rate": 9.066470031475236e-06,
+      "loss": 0.4762,
+      "step": 2279
+    },
+    {
+      "epoch": 0.22189781021897811,
+      "grad_norm": 1.437724469115563,
+      "learning_rate": 9.065552738856745e-06,
+      "loss": 0.3687,
+      "step": 2280
+    },
+    {
+      "epoch": 0.22199513381995134,
+      "grad_norm": 1.2431258010669888,
+      "learning_rate": 9.06463504224178e-06,
+      "loss": 0.3854,
+      "step": 2281
+    },
+    {
+      "epoch": 0.22209245742092457,
+      "grad_norm": 1.362042407967867,
+      "learning_rate": 9.063716941721534e-06,
+      "loss": 0.3981,
+      "step": 2282
+    },
+    {
+      "epoch": 0.22218978102189782,
+      "grad_norm": 1.3260780267557537,
+      "learning_rate": 9.062798437387236e-06,
+      "loss": 0.4304,
+      "step": 2283
+    },
+    {
+      "epoch": 0.22228710462287105,
+      "grad_norm": 1.2009742636293355,
+      "learning_rate": 9.06187952933016e-06,
+      "loss": 0.3441,
+      "step": 2284
+    },
+    {
+      "epoch": 0.22238442822384427,
+      "grad_norm": 1.7089934430562992,
+      "learning_rate": 9.060960217641618e-06,
+      "loss": 0.3488,
+      "step": 2285
+    },
+    {
+      "epoch": 0.22248175182481753,
+      "grad_norm": 1.3539106224768682,
+      "learning_rate": 9.060040502412965e-06,
+      "loss": 0.3617,
+      "step": 2286
+    },
+    {
+      "epoch": 0.22257907542579075,
+      "grad_norm": 1.3952537396094973,
+      "learning_rate": 9.05912038373559e-06,
+      "loss": 0.4507,
+      "step": 2287
+    },
+    {
+      "epoch": 0.22267639902676398,
+      "grad_norm": 1.201207552744405,
+      "learning_rate": 9.058199861700928e-06,
+      "loss": 0.3074,
+      "step": 2288
+    },
+    {
+      "epoch": 0.22277372262773723,
+      "grad_norm": 1.1918182161083974,
+      "learning_rate": 9.057278936400453e-06,
+      "loss": 0.3713,
+      "step": 2289
+    },
+    {
+      "epoch": 0.22287104622871046,
+      "grad_norm": 1.5864015097741249,
+      "learning_rate": 9.056357607925674e-06,
+      "loss": 0.4651,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2229683698296837,
+      "grad_norm": 1.0855034708664277,
+      "learning_rate": 9.055435876368148e-06,
+      "loss": 0.2361,
+      "step": 2291
+    },
+    {
+      "epoch": 0.22306569343065694,
+      "grad_norm": 1.1945153364440069,
+      "learning_rate": 9.054513741819466e-06,
+      "loss": 0.2803,
+      "step": 2292
+    },
+    {
+      "epoch": 0.22316301703163016,
+      "grad_norm": 1.3734264039165323,
+      "learning_rate": 9.053591204371262e-06,
+      "loss": 0.3709,
+      "step": 2293
+    },
+    {
+      "epoch": 0.22326034063260342,
+      "grad_norm": 1.662571628719731,
+      "learning_rate": 9.052668264115206e-06,
+      "loss": 0.6615,
+      "step": 2294
+    },
+    {
+      "epoch": 0.22335766423357664,
+      "grad_norm": 1.4371203045482563,
+      "learning_rate": 9.051744921143015e-06,
+      "loss": 0.4082,
+      "step": 2295
+    },
+    {
+      "epoch": 0.22345498783454987,
+      "grad_norm": 1.5571182647752952,
+      "learning_rate": 9.050821175546442e-06,
+      "loss": 0.5338,
+      "step": 2296
+    },
+    {
+      "epoch": 0.22355231143552312,
+      "grad_norm": 1.4022335338581293,
+      "learning_rate": 9.049897027417277e-06,
+      "loss": 0.3933,
+      "step": 2297
+    },
+    {
+      "epoch": 0.22364963503649635,
+      "grad_norm": 1.2815006290096387,
+      "learning_rate": 9.048972476847356e-06,
+      "loss": 0.4662,
+      "step": 2298
+    },
+    {
+      "epoch": 0.22374695863746957,
+      "grad_norm": 1.4344706750679865,
+      "learning_rate": 9.04804752392855e-06,
+      "loss": 0.4422,
+      "step": 2299
+    },
+    {
+      "epoch": 0.22384428223844283,
+      "grad_norm": 1.2984999163116793,
+      "learning_rate": 9.047122168752775e-06,
+      "loss": 0.3659,
+      "step": 2300
+    },
+    {
+      "epoch": 0.22394160583941605,
+      "grad_norm": 1.1587669196843096,
+      "learning_rate": 9.046196411411982e-06,
+      "loss": 0.2974,
+      "step": 2301
+    },
+    {
+      "epoch": 0.2240389294403893,
+      "grad_norm": 2.322228254141064,
+      "learning_rate": 9.045270251998166e-06,
+      "loss": 0.5667,
+      "step": 2302
+    },
+    {
+      "epoch": 0.22413625304136253,
+      "grad_norm": 1.5137300738559605,
+      "learning_rate": 9.044343690603358e-06,
+      "loss": 0.3889,
+      "step": 2303
+    },
+    {
+      "epoch": 0.22423357664233576,
+      "grad_norm": 1.472679239189759,
+      "learning_rate": 9.04341672731963e-06,
+      "loss": 0.4875,
+      "step": 2304
+    },
+    {
+      "epoch": 0.224330900243309,
+      "grad_norm": 1.391957619608358,
+      "learning_rate": 9.042489362239097e-06,
+      "loss": 0.4513,
+      "step": 2305
+    },
+    {
+      "epoch": 0.22442822384428224,
+      "grad_norm": 1.5752423841676473,
+      "learning_rate": 9.041561595453914e-06,
+      "loss": 0.6021,
+      "step": 2306
+    },
+    {
+      "epoch": 0.22452554744525546,
+      "grad_norm": 1.340696458312585,
+      "learning_rate": 9.040633427056268e-06,
+      "loss": 0.36,
+      "step": 2307
+    },
+    {
+      "epoch": 0.22462287104622872,
+      "grad_norm": 1.319309191993897,
+      "learning_rate": 9.039704857138396e-06,
+      "loss": 0.2632,
+      "step": 2308
+    },
+    {
+      "epoch": 0.22472019464720194,
+      "grad_norm": 1.3567748798839634,
+      "learning_rate": 9.03877588579257e-06,
+      "loss": 0.4085,
+      "step": 2309
+    },
+    {
+      "epoch": 0.22481751824817517,
+      "grad_norm": 1.7234931003044007,
+      "learning_rate": 9.0378465131111e-06,
+      "loss": 0.5366,
+      "step": 2310
+    },
+    {
+      "epoch": 0.22491484184914842,
+      "grad_norm": 1.3431964443797024,
+      "learning_rate": 9.036916739186341e-06,
+      "loss": 0.3406,
+      "step": 2311
+    },
+    {
+      "epoch": 0.22501216545012165,
+      "grad_norm": 1.6143507102825565,
+      "learning_rate": 9.035986564110685e-06,
+      "loss": 0.6322,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2251094890510949,
+      "grad_norm": 1.421713348254314,
+      "learning_rate": 9.035055987976563e-06,
+      "loss": 0.3963,
+      "step": 2313
+    },
+    {
+      "epoch": 0.22520681265206813,
+      "grad_norm": 1.5860325075452377,
+      "learning_rate": 9.034125010876447e-06,
+      "loss": 0.4722,
+      "step": 2314
+    },
+    {
+      "epoch": 0.22530413625304135,
+      "grad_norm": 1.633700480684755,
+      "learning_rate": 9.03319363290285e-06,
+      "loss": 0.2649,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2254014598540146,
+      "grad_norm": 1.5598775600409591,
+      "learning_rate": 9.03226185414832e-06,
+      "loss": 0.4778,
+      "step": 2316
+    },
+    {
+      "epoch": 0.22549878345498783,
+      "grad_norm": 1.4413798673536165,
+      "learning_rate": 9.031329674705455e-06,
+      "loss": 0.3182,
+      "step": 2317
+    },
+    {
+      "epoch": 0.22559610705596106,
+      "grad_norm": 1.437989358950148,
+      "learning_rate": 9.03039709466688e-06,
+      "loss": 0.4297,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2256934306569343,
+      "grad_norm": 1.3355568683760275,
+      "learning_rate": 9.029464114125267e-06,
+      "loss": 0.3393,
+      "step": 2319
+    },
+    {
+      "epoch": 0.22579075425790754,
+      "grad_norm": 1.353161962413978,
+      "learning_rate": 9.028530733173332e-06,
+      "loss": 0.3362,
+      "step": 2320
+    },
+    {
+      "epoch": 0.22588807785888076,
+      "grad_norm": 1.1699742479017108,
+      "learning_rate": 9.027596951903819e-06,
+      "loss": 0.3674,
+      "step": 2321
+    },
+    {
+      "epoch": 0.22598540145985402,
+      "grad_norm": 1.1235278882417843,
+      "learning_rate": 9.026662770409524e-06,
+      "loss": 0.3209,
+      "step": 2322
+    },
+    {
+      "epoch": 0.22608272506082724,
+      "grad_norm": 1.4951135995374567,
+      "learning_rate": 9.025728188783273e-06,
+      "loss": 0.4297,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2261800486618005,
+      "grad_norm": 1.3046514997255336,
+      "learning_rate": 9.024793207117937e-06,
+      "loss": 0.3765,
+      "step": 2324
+    },
+    {
+      "epoch": 0.22627737226277372,
+      "grad_norm": 1.3346554142143854,
+      "learning_rate": 9.023857825506426e-06,
+      "loss": 0.5228,
+      "step": 2325
+    },
+    {
+      "epoch": 0.22637469586374695,
+      "grad_norm": 1.4309619163867682,
+      "learning_rate": 9.022922044041691e-06,
+      "loss": 0.4605,
+      "step": 2326
+    },
+    {
+      "epoch": 0.2264720194647202,
+      "grad_norm": 1.5152634651556307,
+      "learning_rate": 9.021985862816718e-06,
+      "loss": 0.5553,
+      "step": 2327
+    },
+    {
+      "epoch": 0.22656934306569343,
+      "grad_norm": 1.3885182055556289,
+      "learning_rate": 9.02104928192454e-06,
+      "loss": 0.4831,
+      "step": 2328
+    },
+    {
+      "epoch": 0.22666666666666666,
+      "grad_norm": 1.2729317064328092,
+      "learning_rate": 9.020112301458221e-06,
+      "loss": 0.4314,
+      "step": 2329
+    },
+    {
+      "epoch": 0.2267639902676399,
+      "grad_norm": 0.9679503678492228,
+      "learning_rate": 9.019174921510874e-06,
+      "loss": 0.1925,
+      "step": 2330
+    },
+    {
+      "epoch": 0.22686131386861313,
+      "grad_norm": 1.4513146393120597,
+      "learning_rate": 9.018237142175643e-06,
+      "loss": 0.5487,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2269586374695864,
+      "grad_norm": 1.5377065039176208,
+      "learning_rate": 9.017298963545718e-06,
+      "loss": 0.4063,
+      "step": 2332
+    },
+    {
+      "epoch": 0.22705596107055961,
+      "grad_norm": 1.0180180453516632,
+      "learning_rate": 9.016360385714324e-06,
+      "loss": 0.2101,
+      "step": 2333
+    },
+    {
+      "epoch": 0.22715328467153284,
+      "grad_norm": 1.3145676629552665,
+      "learning_rate": 9.015421408774732e-06,
+      "loss": 0.4575,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2272506082725061,
+      "grad_norm": 1.3213351651174845,
+      "learning_rate": 9.014482032820247e-06,
+      "loss": 0.3924,
+      "step": 2335
+    },
+    {
+      "epoch": 0.22734793187347932,
+      "grad_norm": 1.9370834148842127,
+      "learning_rate": 9.013542257944212e-06,
+      "loss": 0.4332,
+      "step": 2336
+    },
+    {
+      "epoch": 0.22744525547445255,
+      "grad_norm": 1.4754695985325648,
+      "learning_rate": 9.012602084240018e-06,
+      "loss": 0.4014,
+      "step": 2337
+    },
+    {
+      "epoch": 0.2275425790754258,
+      "grad_norm": 1.1124893316550342,
+      "learning_rate": 9.011661511801088e-06,
+      "loss": 0.2957,
+      "step": 2338
+    },
+    {
+      "epoch": 0.22763990267639903,
+      "grad_norm": 1.2537185195667433,
+      "learning_rate": 9.010720540720888e-06,
+      "loss": 0.3004,
+      "step": 2339
+    },
+    {
+      "epoch": 0.22773722627737225,
+      "grad_norm": 1.4597689601256807,
+      "learning_rate": 9.009779171092923e-06,
+      "loss": 0.2555,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2278345498783455,
+      "grad_norm": 1.4737791439989423,
+      "learning_rate": 9.008837403010736e-06,
+      "loss": 0.5355,
+      "step": 2341
+    },
+    {
+      "epoch": 0.22793187347931873,
+      "grad_norm": 1.3795639069131398,
+      "learning_rate": 9.007895236567913e-06,
+      "loss": 0.3961,
+      "step": 2342
+    },
+    {
+      "epoch": 0.22802919708029198,
+      "grad_norm": 1.6364796903185053,
+      "learning_rate": 9.006952671858078e-06,
+      "loss": 0.444,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2281265206812652,
+      "grad_norm": 1.1964346909925698,
+      "learning_rate": 9.006009708974892e-06,
+      "loss": 0.3297,
+      "step": 2344
+    },
+    {
+      "epoch": 0.22822384428223844,
+      "grad_norm": 1.343808771666808,
+      "learning_rate": 9.00506634801206e-06,
+      "loss": 0.4537,
+      "step": 2345
+    },
+    {
+      "epoch": 0.2283211678832117,
+      "grad_norm": 1.4003110727355261,
+      "learning_rate": 9.004122589063323e-06,
+      "loss": 0.3883,
+      "step": 2346
+    },
+    {
+      "epoch": 0.22841849148418492,
+      "grad_norm": 1.2435101838594087,
+      "learning_rate": 9.003178432222462e-06,
+      "loss": 0.4238,
+      "step": 2347
+    },
+    {
+      "epoch": 0.22851581508515814,
+      "grad_norm": 1.324643227390155,
+      "learning_rate": 9.0022338775833e-06,
+      "loss": 0.4139,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2286131386861314,
+      "grad_norm": 1.7692069120616638,
+      "learning_rate": 9.001288925239698e-06,
+      "loss": 0.4719,
+      "step": 2349
+    },
+    {
+      "epoch": 0.22871046228710462,
+      "grad_norm": 1.223562422765287,
+      "learning_rate": 9.000343575285555e-06,
+      "loss": 0.3256,
+      "step": 2350
+    },
+    {
+      "epoch": 0.22880778588807785,
+      "grad_norm": 1.3407025045830592,
+      "learning_rate": 8.999397827814812e-06,
+      "loss": 0.3788,
+      "step": 2351
+    },
+    {
+      "epoch": 0.2289051094890511,
+      "grad_norm": 1.5281139100341292,
+      "learning_rate": 8.99845168292145e-06,
+      "loss": 0.5565,
+      "step": 2352
+    },
+    {
+      "epoch": 0.22900243309002433,
+      "grad_norm": 1.560155712083658,
+      "learning_rate": 8.997505140699488e-06,
+      "loss": 0.4957,
+      "step": 2353
+    },
+    {
+      "epoch": 0.22909975669099758,
+      "grad_norm": 1.290422773797366,
+      "learning_rate": 8.996558201242981e-06,
+      "loss": 0.4011,
+      "step": 2354
+    },
+    {
+      "epoch": 0.2291970802919708,
+      "grad_norm": 1.2847680894150124,
+      "learning_rate": 8.99561086464603e-06,
+      "loss": 0.4419,
+      "step": 2355
+    },
+    {
+      "epoch": 0.22929440389294403,
+      "grad_norm": 1.4625413220428547,
+      "learning_rate": 8.99466313100277e-06,
+      "loss": 0.2511,
+      "step": 2356
+    },
+    {
+      "epoch": 0.22939172749391729,
+      "grad_norm": 1.2882840667194135,
+      "learning_rate": 8.99371500040738e-06,
+      "loss": 0.3992,
+      "step": 2357
+    },
+    {
+      "epoch": 0.2294890510948905,
+      "grad_norm": 1.1997126453782205,
+      "learning_rate": 8.992766472954077e-06,
+      "loss": 0.2639,
+      "step": 2358
+    },
+    {
+      "epoch": 0.22958637469586374,
+      "grad_norm": 1.6688893120724655,
+      "learning_rate": 8.991817548737114e-06,
+      "loss": 0.3103,
+      "step": 2359
+    },
+    {
+      "epoch": 0.229683698296837,
+      "grad_norm": 1.4031771252981649,
+      "learning_rate": 8.990868227850788e-06,
+      "loss": 0.4245,
+      "step": 2360
+    },
+    {
+      "epoch": 0.22978102189781022,
+      "grad_norm": 1.4825462721346627,
+      "learning_rate": 8.989918510389432e-06,
+      "loss": 0.3973,
+      "step": 2361
+    },
+    {
+      "epoch": 0.22987834549878344,
+      "grad_norm": 1.7756990641125774,
+      "learning_rate": 8.988968396447424e-06,
+      "loss": 0.6091,
+      "step": 2362
+    },
+    {
+      "epoch": 0.2299756690997567,
+      "grad_norm": 1.5519381803018173,
+      "learning_rate": 8.988017886119172e-06,
+      "loss": 0.5849,
+      "step": 2363
+    },
+    {
+      "epoch": 0.23007299270072992,
+      "grad_norm": 1.5288537407748173,
+      "learning_rate": 8.987066979499133e-06,
+      "loss": 0.594,
+      "step": 2364
+    },
+    {
+      "epoch": 0.23017031630170318,
+      "grad_norm": 1.2519254160654887,
+      "learning_rate": 8.986115676681797e-06,
+      "loss": 0.3781,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2302676399026764,
+      "grad_norm": 1.2118409754918265,
+      "learning_rate": 8.985163977761697e-06,
+      "loss": 0.3761,
+      "step": 2366
+    },
+    {
+      "epoch": 0.23036496350364963,
+      "grad_norm": 1.3123505825187787,
+      "learning_rate": 8.984211882833402e-06,
+      "loss": 0.405,
+      "step": 2367
+    },
+    {
+      "epoch": 0.23046228710462288,
+      "grad_norm": 1.6027642184293107,
+      "learning_rate": 8.983259391991524e-06,
+      "loss": 0.597,
+      "step": 2368
+    },
+    {
+      "epoch": 0.2305596107055961,
+      "grad_norm": 1.3646497443348367,
+      "learning_rate": 8.982306505330712e-06,
+      "loss": 0.4036,
+      "step": 2369
+    },
+    {
+      "epoch": 0.23065693430656933,
+      "grad_norm": 1.2894115553392402,
+      "learning_rate": 8.981353222945653e-06,
+      "loss": 0.2778,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2307542579075426,
+      "grad_norm": 1.27883786418869,
+      "learning_rate": 8.98039954493108e-06,
+      "loss": 0.3803,
+      "step": 2371
+    },
+    {
+      "epoch": 0.2308515815085158,
+      "grad_norm": 1.5863647637061415,
+      "learning_rate": 8.979445471381755e-06,
+      "loss": 0.4716,
+      "step": 2372
+    },
+    {
+      "epoch": 0.23094890510948904,
+      "grad_norm": 1.1874137646332688,
+      "learning_rate": 8.97849100239249e-06,
+      "loss": 0.2846,
+      "step": 2373
+    },
+    {
+      "epoch": 0.2310462287104623,
+      "grad_norm": 1.6936318641369774,
+      "learning_rate": 8.977536138058126e-06,
+      "loss": 0.3418,
+      "step": 2374
+    },
+    {
+      "epoch": 0.23114355231143552,
+      "grad_norm": 1.0526167143851337,
+      "learning_rate": 8.976580878473553e-06,
+      "loss": 0.258,
+      "step": 2375
+    },
+    {
+      "epoch": 0.23124087591240877,
+      "grad_norm": 1.753799809070063,
+      "learning_rate": 8.975625223733693e-06,
+      "loss": 0.4764,
+      "step": 2376
+    },
+    {
+      "epoch": 0.231338199513382,
+      "grad_norm": 1.3814482775299988,
+      "learning_rate": 8.97466917393351e-06,
+      "loss": 0.3811,
+      "step": 2377
+    },
+    {
+      "epoch": 0.23143552311435522,
+      "grad_norm": 1.575424754678499,
+      "learning_rate": 8.97371272916801e-06,
+      "loss": 0.5028,
+      "step": 2378
+    },
+    {
+      "epoch": 0.23153284671532848,
+      "grad_norm": 1.5163540217481704,
+      "learning_rate": 8.972755889532234e-06,
+      "loss": 0.4055,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2316301703163017,
+      "grad_norm": 1.1877796947964157,
+      "learning_rate": 8.971798655121264e-06,
+      "loss": 0.2978,
+      "step": 2380
+    },
+    {
+      "epoch": 0.23172749391727493,
+      "grad_norm": 1.6274909221671408,
+      "learning_rate": 8.970841026030218e-06,
+      "loss": 0.4319,
+      "step": 2381
+    },
+    {
+      "epoch": 0.23182481751824818,
+      "grad_norm": 1.413480143472021,
+      "learning_rate": 8.969883002354259e-06,
+      "loss": 0.4015,
+      "step": 2382
+    },
+    {
+      "epoch": 0.2319221411192214,
+      "grad_norm": 1.451327617189514,
+      "learning_rate": 8.968924584188587e-06,
+      "loss": 0.5107,
+      "step": 2383
+    },
+    {
+      "epoch": 0.23201946472019463,
+      "grad_norm": 1.4288160659587352,
+      "learning_rate": 8.96796577162844e-06,
+      "loss": 0.369,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2321167883211679,
+      "grad_norm": 1.6469132304956866,
+      "learning_rate": 8.967006564769094e-06,
+      "loss": 0.5982,
+      "step": 2385
+    },
+    {
+      "epoch": 0.23221411192214111,
+      "grad_norm": 1.4887239693800984,
+      "learning_rate": 8.966046963705869e-06,
+      "loss": 0.4967,
+      "step": 2386
+    },
+    {
+      "epoch": 0.23231143552311437,
+      "grad_norm": 1.2469481884120308,
+      "learning_rate": 8.965086968534116e-06,
+      "loss": 0.4022,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2324087591240876,
+      "grad_norm": 2.6320603198934047,
+      "learning_rate": 8.964126579349237e-06,
+      "loss": 0.2489,
+      "step": 2388
+    },
+    {
+      "epoch": 0.23250608272506082,
+      "grad_norm": 1.2339093742509784,
+      "learning_rate": 8.963165796246663e-06,
+      "loss": 0.3694,
+      "step": 2389
+    },
+    {
+      "epoch": 0.23260340632603407,
+      "grad_norm": 1.4634162966788549,
+      "learning_rate": 8.962204619321866e-06,
+      "loss": 0.5646,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2327007299270073,
+      "grad_norm": 1.2919651066139786,
+      "learning_rate": 8.961243048670363e-06,
+      "loss": 0.3833,
+      "step": 2391
+    },
+    {
+      "epoch": 0.23279805352798053,
+      "grad_norm": 1.5273773111622013,
+      "learning_rate": 8.960281084387701e-06,
+      "loss": 0.5724,
+      "step": 2392
+    },
+    {
+      "epoch": 0.23289537712895378,
+      "grad_norm": 1.4704498843019616,
+      "learning_rate": 8.959318726569475e-06,
+      "loss": 0.5232,
+      "step": 2393
+    },
+    {
+      "epoch": 0.232992700729927,
+      "grad_norm": 1.52947786509823,
+      "learning_rate": 8.958355975311314e-06,
+      "loss": 0.5014,
+      "step": 2394
+    },
+    {
+      "epoch": 0.23309002433090023,
+      "grad_norm": 1.457234959002331,
+      "learning_rate": 8.957392830708886e-06,
+      "loss": 0.5401,
+      "step": 2395
+    },
+    {
+      "epoch": 0.23318734793187348,
+      "grad_norm": 1.5878948291380384,
+      "learning_rate": 8.9564292928579e-06,
+      "loss": 0.4481,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2332846715328467,
+      "grad_norm": 1.3353181262068508,
+      "learning_rate": 8.955465361854103e-06,
+      "loss": 0.3668,
+      "step": 2397
+    },
+    {
+      "epoch": 0.23338199513381996,
+      "grad_norm": 2.023729457927684,
+      "learning_rate": 8.954501037793282e-06,
+      "loss": 0.256,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2334793187347932,
+      "grad_norm": 1.3501136378744423,
+      "learning_rate": 8.953536320771264e-06,
+      "loss": 0.4288,
+      "step": 2399
+    },
+    {
+      "epoch": 0.23357664233576642,
+      "grad_norm": 0.9695156209886321,
+      "learning_rate": 8.95257121088391e-06,
+      "loss": 0.3313,
+      "step": 2400
+    },
+    {
+      "epoch": 0.23367396593673967,
+      "grad_norm": 1.6268089203048999,
+      "learning_rate": 8.951605708227125e-06,
+      "loss": 0.5031,
+      "step": 2401
+    },
+    {
+      "epoch": 0.2337712895377129,
+      "grad_norm": 1.3327356528771297,
+      "learning_rate": 8.950639812896852e-06,
+      "loss": 0.352,
+      "step": 2402
+    },
+    {
+      "epoch": 0.23386861313868612,
+      "grad_norm": 1.646158604731562,
+      "learning_rate": 8.949673524989074e-06,
+      "loss": 0.6143,
+      "step": 2403
+    },
+    {
+      "epoch": 0.23396593673965937,
+      "grad_norm": 1.4459398712277267,
+      "learning_rate": 8.948706844599809e-06,
+      "loss": 0.301,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2340632603406326,
+      "grad_norm": 1.242464142709881,
+      "learning_rate": 8.947739771825118e-06,
+      "loss": 0.3867,
+      "step": 2405
+    },
+    {
+      "epoch": 0.23416058394160583,
+      "grad_norm": 1.283369590610404,
+      "learning_rate": 8.946772306761099e-06,
+      "loss": 0.3396,
+      "step": 2406
+    },
+    {
+      "epoch": 0.23425790754257908,
+      "grad_norm": 1.659051576981879,
+      "learning_rate": 8.94580444950389e-06,
+      "loss": 0.2985,
+      "step": 2407
+    },
+    {
+      "epoch": 0.2343552311435523,
+      "grad_norm": 1.5811557183787177,
+      "learning_rate": 8.944836200149669e-06,
+      "loss": 0.5412,
+      "step": 2408
+    },
+    {
+      "epoch": 0.23445255474452556,
+      "grad_norm": 1.5937284580345608,
+      "learning_rate": 8.943867558794648e-06,
+      "loss": 0.4562,
+      "step": 2409
+    },
+    {
+      "epoch": 0.23454987834549879,
+      "grad_norm": 1.179539450140548,
+      "learning_rate": 8.942898525535085e-06,
+      "loss": 0.2436,
+      "step": 2410
+    },
+    {
+      "epoch": 0.234647201946472,
+      "grad_norm": 1.2115140465312926,
+      "learning_rate": 8.941929100467272e-06,
+      "loss": 0.325,
+      "step": 2411
+    },
+    {
+      "epoch": 0.23474452554744527,
+      "grad_norm": 1.3228525862104779,
+      "learning_rate": 8.94095928368754e-06,
+      "loss": 0.4001,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2348418491484185,
+      "grad_norm": 1.5093562470528878,
+      "learning_rate": 8.939989075292263e-06,
+      "loss": 0.3554,
+      "step": 2413
+    },
+    {
+      "epoch": 0.23493917274939172,
+      "grad_norm": 1.629660086616085,
+      "learning_rate": 8.93901847537785e-06,
+      "loss": 0.6349,
+      "step": 2414
+    },
+    {
+      "epoch": 0.23503649635036497,
+      "grad_norm": 1.0826348229158524,
+      "learning_rate": 8.938047484040749e-06,
+      "loss": 0.2681,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2351338199513382,
+      "grad_norm": 1.2841520241198179,
+      "learning_rate": 8.93707610137745e-06,
+      "loss": 0.4081,
+      "step": 2416
+    },
+    {
+      "epoch": 0.23523114355231142,
+      "grad_norm": 1.913465881785096,
+      "learning_rate": 8.936104327484479e-06,
+      "loss": 0.7043,
+      "step": 2417
+    },
+    {
+      "epoch": 0.23532846715328468,
+      "grad_norm": 1.386306701477425,
+      "learning_rate": 8.935132162458401e-06,
+      "loss": 0.341,
+      "step": 2418
+    },
+    {
+      "epoch": 0.2354257907542579,
+      "grad_norm": 1.1278547518059516,
+      "learning_rate": 8.934159606395821e-06,
+      "loss": 0.3151,
+      "step": 2419
+    },
+    {
+      "epoch": 0.23552311435523116,
+      "grad_norm": 1.5540265542588236,
+      "learning_rate": 8.933186659393384e-06,
+      "loss": 0.6514,
+      "step": 2420
+    },
+    {
+      "epoch": 0.23562043795620438,
+      "grad_norm": 1.278787339635804,
+      "learning_rate": 8.932213321547769e-06,
+      "loss": 0.3423,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2357177615571776,
+      "grad_norm": 1.2885094583361822,
+      "learning_rate": 8.931239592955701e-06,
+      "loss": 0.2958,
+      "step": 2422
+    },
+    {
+      "epoch": 0.23581508515815086,
+      "grad_norm": 1.5181901598500283,
+      "learning_rate": 8.930265473713939e-06,
+      "loss": 0.4212,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2359124087591241,
+      "grad_norm": 1.2136160482551297,
+      "learning_rate": 8.92929096391928e-06,
+      "loss": 0.3982,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2360097323600973,
+      "grad_norm": 1.5487072814518004,
+      "learning_rate": 8.928316063668562e-06,
+      "loss": 0.5676,
+      "step": 2425
+    },
+    {
+      "epoch": 0.23610705596107057,
+      "grad_norm": 1.430432818475582,
+      "learning_rate": 8.927340773058664e-06,
+      "loss": 0.4735,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2362043795620438,
+      "grad_norm": 1.4586841524588252,
+      "learning_rate": 8.926365092186498e-06,
+      "loss": 0.5637,
+      "step": 2427
+    },
+    {
+      "epoch": 0.23630170316301702,
+      "grad_norm": 1.5364014523424565,
+      "learning_rate": 8.92538902114902e-06,
+      "loss": 0.4783,
+      "step": 2428
+    },
+    {
+      "epoch": 0.23639902676399027,
+      "grad_norm": 1.3896600182614345,
+      "learning_rate": 8.924412560043223e-06,
+      "loss": 0.3748,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2364963503649635,
+      "grad_norm": 1.304447540327908,
+      "learning_rate": 8.923435708966135e-06,
+      "loss": 0.3373,
+      "step": 2430
+    },
+    {
+      "epoch": 0.23659367396593675,
+      "grad_norm": 1.3383082719469825,
+      "learning_rate": 8.922458468014833e-06,
+      "loss": 0.3089,
+      "step": 2431
+    },
+    {
+      "epoch": 0.23669099756690998,
+      "grad_norm": 1.4376693294142868,
+      "learning_rate": 8.921480837286418e-06,
+      "loss": 0.2665,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2367883211678832,
+      "grad_norm": 1.3948368197200884,
+      "learning_rate": 8.920502816878045e-06,
+      "loss": 0.4349,
+      "step": 2433
+    },
+    {
+      "epoch": 0.23688564476885646,
+      "grad_norm": 1.5583938814663865,
+      "learning_rate": 8.919524406886897e-06,
+      "loss": 0.4528,
+      "step": 2434
+    },
+    {
+      "epoch": 0.23698296836982968,
+      "grad_norm": 1.455016515054737,
+      "learning_rate": 8.918545607410199e-06,
+      "loss": 0.416,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2370802919708029,
+      "grad_norm": 1.5707414335423742,
+      "learning_rate": 8.917566418545215e-06,
+      "loss": 0.4269,
+      "step": 2436
+    },
+    {
+      "epoch": 0.23717761557177616,
+      "grad_norm": 1.6214497738286784,
+      "learning_rate": 8.916586840389248e-06,
+      "loss": 0.5531,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2372749391727494,
+      "grad_norm": 1.5231468510302828,
+      "learning_rate": 8.91560687303964e-06,
+      "loss": 0.5464,
+      "step": 2438
+    },
+    {
+      "epoch": 0.23737226277372261,
+      "grad_norm": 1.5631657517225734,
+      "learning_rate": 8.91462651659377e-06,
+      "loss": 0.4098,
+      "step": 2439
+    },
+    {
+      "epoch": 0.23746958637469587,
+      "grad_norm": 1.5003582208774642,
+      "learning_rate": 8.913645771149058e-06,
+      "loss": 0.342,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2375669099756691,
+      "grad_norm": 1.2703591332316027,
+      "learning_rate": 8.91266463680296e-06,
+      "loss": 0.3195,
+      "step": 2441
+    },
+    {
+      "epoch": 0.23766423357664235,
+      "grad_norm": 1.3910967851640175,
+      "learning_rate": 8.91168311365297e-06,
+      "loss": 0.334,
+      "step": 2442
+    },
+    {
+      "epoch": 0.23776155717761557,
+      "grad_norm": 1.5001053773105038,
+      "learning_rate": 8.910701201796625e-06,
+      "loss": 0.4665,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2378588807785888,
+      "grad_norm": 1.6142849143926903,
+      "learning_rate": 8.9097189013315e-06,
+      "loss": 0.5276,
+      "step": 2444
+    },
+    {
+      "epoch": 0.23795620437956205,
+      "grad_norm": 1.2059866820401877,
+      "learning_rate": 8.908736212355202e-06,
+      "loss": 0.2936,
+      "step": 2445
+    },
+    {
+      "epoch": 0.23805352798053528,
+      "grad_norm": 1.4496663268694052,
+      "learning_rate": 8.907753134965387e-06,
+      "loss": 0.475,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2381508515815085,
+      "grad_norm": 1.4184456855989886,
+      "learning_rate": 8.90676966925974e-06,
+      "loss": 0.4477,
+      "step": 2447
+    },
+    {
+      "epoch": 0.23824817518248176,
+      "grad_norm": 1.7126804340284862,
+      "learning_rate": 8.90578581533599e-06,
+      "loss": 0.6392,
+      "step": 2448
+    },
+    {
+      "epoch": 0.23834549878345498,
+      "grad_norm": 1.6085356958926766,
+      "learning_rate": 8.904801573291901e-06,
+      "loss": 0.4428,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2384428223844282,
+      "grad_norm": 1.1724096477321129,
+      "learning_rate": 8.903816943225281e-06,
+      "loss": 0.23,
+      "step": 2450
+    },
+    {
+      "epoch": 0.23854014598540146,
+      "grad_norm": 1.2849397331978023,
+      "learning_rate": 8.902831925233972e-06,
+      "loss": 0.4315,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2386374695863747,
+      "grad_norm": 1.3479724015628292,
+      "learning_rate": 8.901846519415856e-06,
+      "loss": 0.4528,
+      "step": 2452
+    },
+    {
+      "epoch": 0.23873479318734794,
+      "grad_norm": 1.5241447958707557,
+      "learning_rate": 8.900860725868852e-06,
+      "loss": 0.5638,
+      "step": 2453
+    },
+    {
+      "epoch": 0.23883211678832117,
+      "grad_norm": 1.3008951589753057,
+      "learning_rate": 8.899874544690921e-06,
+      "loss": 0.4364,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2389294403892944,
+      "grad_norm": 1.3889516127516133,
+      "learning_rate": 8.89888797598006e-06,
+      "loss": 0.5968,
+      "step": 2455
+    },
+    {
+      "epoch": 0.23902676399026765,
+      "grad_norm": 1.3382384356293548,
+      "learning_rate": 8.8979010198343e-06,
+      "loss": 0.3423,
+      "step": 2456
+    },
+    {
+      "epoch": 0.23912408759124087,
+      "grad_norm": 1.3927455122024084,
+      "learning_rate": 8.896913676351726e-06,
+      "loss": 0.5291,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2392214111922141,
+      "grad_norm": 1.3654704619725508,
+      "learning_rate": 8.895925945630441e-06,
+      "loss": 0.3224,
+      "step": 2458
+    },
+    {
+      "epoch": 0.23931873479318735,
+      "grad_norm": 2.420859240745107,
+      "learning_rate": 8.8949378277686e-06,
+      "loss": 0.4526,
+      "step": 2459
+    },
+    {
+      "epoch": 0.23941605839416058,
+      "grad_norm": 1.279171164356654,
+      "learning_rate": 8.893949322864394e-06,
+      "loss": 0.3452,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2395133819951338,
+      "grad_norm": 1.4336845514712926,
+      "learning_rate": 8.89296043101605e-06,
+      "loss": 0.3891,
+      "step": 2461
+    },
+    {
+      "epoch": 0.23961070559610706,
+      "grad_norm": 1.2391493008048138,
+      "learning_rate": 8.891971152321836e-06,
+      "loss": 0.5135,
+      "step": 2462
+    },
+    {
+      "epoch": 0.23970802919708029,
+      "grad_norm": 1.2398633987802397,
+      "learning_rate": 8.890981486880057e-06,
+      "loss": 0.2688,
+      "step": 2463
+    },
+    {
+      "epoch": 0.23980535279805354,
+      "grad_norm": 1.1975725536626207,
+      "learning_rate": 8.889991434789054e-06,
+      "loss": 0.4181,
+      "step": 2464
+    },
+    {
+      "epoch": 0.23990267639902677,
+      "grad_norm": 1.5121790458693565,
+      "learning_rate": 8.889000996147213e-06,
+      "loss": 0.667,
+      "step": 2465
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2980809407294283,
+      "learning_rate": 8.888010171052951e-06,
+      "loss": 0.4025,
+      "step": 2466
+    },
+    {
+      "epoch": 0.24009732360097324,
+      "grad_norm": 1.3683247659037883,
+      "learning_rate": 8.887018959604731e-06,
+      "loss": 0.4195,
+      "step": 2467
+    },
+    {
+      "epoch": 0.24019464720194647,
+      "grad_norm": 1.6392091000056277,
+      "learning_rate": 8.886027361901045e-06,
+      "loss": 0.4464,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2402919708029197,
+      "grad_norm": 1.4286158146093557,
+      "learning_rate": 8.885035378040435e-06,
+      "loss": 0.503,
+      "step": 2469
+    },
+    {
+      "epoch": 0.24038929440389295,
+      "grad_norm": 1.6249203295617591,
+      "learning_rate": 8.884043008121468e-06,
+      "loss": 0.5875,
+      "step": 2470
+    },
+    {
+      "epoch": 0.24048661800486618,
+      "grad_norm": 1.316531393288964,
+      "learning_rate": 8.883050252242762e-06,
+      "loss": 0.3225,
+      "step": 2471
+    },
+    {
+      "epoch": 0.24058394160583943,
+      "grad_norm": 1.3738066957140371,
+      "learning_rate": 8.882057110502964e-06,
+      "loss": 0.3863,
+      "step": 2472
+    },
+    {
+      "epoch": 0.24068126520681266,
+      "grad_norm": 1.6149562610100578,
+      "learning_rate": 8.881063583000766e-06,
+      "loss": 0.6899,
+      "step": 2473
+    },
+    {
+      "epoch": 0.24077858880778588,
+      "grad_norm": 1.1978996054498634,
+      "learning_rate": 8.880069669834895e-06,
+      "loss": 0.4647,
+      "step": 2474
+    },
+    {
+      "epoch": 0.24087591240875914,
+      "grad_norm": 1.3737195294986575,
+      "learning_rate": 8.879075371104114e-06,
+      "loss": 0.3404,
+      "step": 2475
+    },
+    {
+      "epoch": 0.24097323600973236,
+      "grad_norm": 1.3242090275500389,
+      "learning_rate": 8.878080686907231e-06,
+      "loss": 0.4923,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2410705596107056,
+      "grad_norm": 1.295191211796917,
+      "learning_rate": 8.877085617343085e-06,
+      "loss": 0.4449,
+      "step": 2477
+    },
+    {
+      "epoch": 0.24116788321167884,
+      "grad_norm": 1.5068542914468723,
+      "learning_rate": 8.87609016251056e-06,
+      "loss": 0.5506,
+      "step": 2478
+    },
+    {
+      "epoch": 0.24126520681265207,
+      "grad_norm": 1.650040845654398,
+      "learning_rate": 8.87509432250857e-06,
+      "loss": 0.5715,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2413625304136253,
+      "grad_norm": 1.5289429392674028,
+      "learning_rate": 8.874098097436078e-06,
+      "loss": 0.5626,
+      "step": 2480
+    },
+    {
+      "epoch": 0.24145985401459855,
+      "grad_norm": 1.3609358059405043,
+      "learning_rate": 8.873101487392078e-06,
+      "loss": 0.4096,
+      "step": 2481
+    },
+    {
+      "epoch": 0.24155717761557177,
+      "grad_norm": 1.5725676631470524,
+      "learning_rate": 8.8721044924756e-06,
+      "loss": 0.6597,
+      "step": 2482
+    },
+    {
+      "epoch": 0.24165450121654503,
+      "grad_norm": 1.094002939677081,
+      "learning_rate": 8.87110711278572e-06,
+      "loss": 0.3206,
+      "step": 2483
+    },
+    {
+      "epoch": 0.24175182481751825,
+      "grad_norm": 1.4551979783640236,
+      "learning_rate": 8.870109348421544e-06,
+      "loss": 0.445,
+      "step": 2484
+    },
+    {
+      "epoch": 0.24184914841849148,
+      "grad_norm": 1.534219781362636,
+      "learning_rate": 8.869111199482227e-06,
+      "loss": 0.6666,
+      "step": 2485
+    },
+    {
+      "epoch": 0.24194647201946473,
+      "grad_norm": 0.9530847884904149,
+      "learning_rate": 8.86811266606695e-06,
+      "loss": 0.2756,
+      "step": 2486
+    },
+    {
+      "epoch": 0.24204379562043796,
+      "grad_norm": 1.4859819247146357,
+      "learning_rate": 8.86711374827494e-06,
+      "loss": 0.4626,
+      "step": 2487
+    },
+    {
+      "epoch": 0.24214111922141118,
+      "grad_norm": 1.5336983239407425,
+      "learning_rate": 8.86611444620546e-06,
+      "loss": 0.5383,
+      "step": 2488
+    },
+    {
+      "epoch": 0.24223844282238444,
+      "grad_norm": 1.4073640437212571,
+      "learning_rate": 8.865114759957812e-06,
+      "loss": 0.4675,
+      "step": 2489
+    },
+    {
+      "epoch": 0.24233576642335766,
+      "grad_norm": 1.562895534043348,
+      "learning_rate": 8.864114689631334e-06,
+      "loss": 0.5641,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2424330900243309,
+      "grad_norm": 1.643145414496213,
+      "learning_rate": 8.863114235325405e-06,
+      "loss": 0.5749,
+      "step": 2491
+    },
+    {
+      "epoch": 0.24253041362530414,
+      "grad_norm": 1.226721686463078,
+      "learning_rate": 8.862113397139437e-06,
+      "loss": 0.3432,
+      "step": 2492
+    },
+    {
+      "epoch": 0.24262773722627737,
+      "grad_norm": 1.2699959241996903,
+      "learning_rate": 8.86111217517289e-06,
+      "loss": 0.4203,
+      "step": 2493
+    },
+    {
+      "epoch": 0.24272506082725062,
+      "grad_norm": 1.4233705808484327,
+      "learning_rate": 8.860110569525253e-06,
+      "loss": 0.2601,
+      "step": 2494
+    },
+    {
+      "epoch": 0.24282238442822385,
+      "grad_norm": 1.3784035260656315,
+      "learning_rate": 8.859108580296055e-06,
+      "loss": 0.4973,
+      "step": 2495
+    },
+    {
+      "epoch": 0.24291970802919707,
+      "grad_norm": 1.2790024746236357,
+      "learning_rate": 8.858106207584864e-06,
+      "loss": 0.4067,
+      "step": 2496
+    },
+    {
+      "epoch": 0.24301703163017033,
+      "grad_norm": 1.4041054798155945,
+      "learning_rate": 8.857103451491292e-06,
+      "loss": 0.5228,
+      "step": 2497
+    },
+    {
+      "epoch": 0.24311435523114355,
+      "grad_norm": 1.6788565066048042,
+      "learning_rate": 8.856100312114975e-06,
+      "loss": 0.7133,
+      "step": 2498
+    },
+    {
+      "epoch": 0.24321167883211678,
+      "grad_norm": 1.2024623978380433,
+      "learning_rate": 8.855096789555602e-06,
+      "loss": 0.2507,
+      "step": 2499
+    },
+    {
+      "epoch": 0.24330900243309003,
+      "grad_norm": 1.3828470689148782,
+      "learning_rate": 8.85409288391289e-06,
+      "loss": 0.3993,
+      "step": 2500
+    },
+    {
+      "epoch": 0.24340632603406326,
+      "grad_norm": 1.427484284296059,
+      "learning_rate": 8.8530885952866e-06,
+      "loss": 0.3926,
+      "step": 2501
+    },
+    {
+      "epoch": 0.24350364963503648,
+      "grad_norm": 1.3193446567792235,
+      "learning_rate": 8.852083923776529e-06,
+      "loss": 0.2152,
+      "step": 2502
+    },
+    {
+      "epoch": 0.24360097323600974,
+      "grad_norm": 1.3297823718570532,
+      "learning_rate": 8.851078869482509e-06,
+      "loss": 0.4772,
+      "step": 2503
+    },
+    {
+      "epoch": 0.24369829683698296,
+      "grad_norm": 1.3044660003313646,
+      "learning_rate": 8.850073432504416e-06,
+      "loss": 0.3589,
+      "step": 2504
+    },
+    {
+      "epoch": 0.24379562043795622,
+      "grad_norm": 1.4488096250914715,
+      "learning_rate": 8.84906761294216e-06,
+      "loss": 0.3261,
+      "step": 2505
+    },
+    {
+      "epoch": 0.24389294403892944,
+      "grad_norm": 1.2778329641523152,
+      "learning_rate": 8.848061410895687e-06,
+      "loss": 0.3047,
+      "step": 2506
+    },
+    {
+      "epoch": 0.24399026763990267,
+      "grad_norm": 1.135638375757245,
+      "learning_rate": 8.847054826464988e-06,
+      "loss": 0.3173,
+      "step": 2507
+    },
+    {
+      "epoch": 0.24408759124087592,
+      "grad_norm": 1.5033745953013864,
+      "learning_rate": 8.846047859750086e-06,
+      "loss": 0.4813,
+      "step": 2508
+    },
+    {
+      "epoch": 0.24418491484184915,
+      "grad_norm": 1.1189501535394493,
+      "learning_rate": 8.845040510851044e-06,
+      "loss": 0.3359,
+      "step": 2509
+    },
+    {
+      "epoch": 0.24428223844282237,
+      "grad_norm": 1.4743455663494507,
+      "learning_rate": 8.844032779867966e-06,
+      "loss": 0.5354,
+      "step": 2510
+    },
+    {
+      "epoch": 0.24437956204379563,
+      "grad_norm": 1.2644405709657818,
+      "learning_rate": 8.843024666900983e-06,
+      "loss": 0.4019,
+      "step": 2511
+    },
+    {
+      "epoch": 0.24447688564476885,
+      "grad_norm": 1.5585250648144962,
+      "learning_rate": 8.84201617205028e-06,
+      "loss": 0.4977,
+      "step": 2512
+    },
+    {
+      "epoch": 0.24457420924574208,
+      "grad_norm": 1.5187811483320863,
+      "learning_rate": 8.841007295416069e-06,
+      "loss": 0.6282,
+      "step": 2513
+    },
+    {
+      "epoch": 0.24467153284671533,
+      "grad_norm": 1.461783750506842,
+      "learning_rate": 8.839998037098601e-06,
+      "loss": 0.6085,
+      "step": 2514
+    },
+    {
+      "epoch": 0.24476885644768856,
+      "grad_norm": 1.4235036556142022,
+      "learning_rate": 8.838988397198167e-06,
+      "loss": 0.5696,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2448661800486618,
+      "grad_norm": 1.6731038078758624,
+      "learning_rate": 8.837978375815097e-06,
+      "loss": 0.5026,
+      "step": 2516
+    },
+    {
+      "epoch": 0.24496350364963504,
+      "grad_norm": 1.2803102163564937,
+      "learning_rate": 8.836967973049757e-06,
+      "loss": 0.2605,
+      "step": 2517
+    },
+    {
+      "epoch": 0.24506082725060827,
+      "grad_norm": 1.2869808613318177,
+      "learning_rate": 8.835957189002551e-06,
+      "loss": 0.3073,
+      "step": 2518
+    },
+    {
+      "epoch": 0.24515815085158152,
+      "grad_norm": 1.4129342483481067,
+      "learning_rate": 8.834946023773921e-06,
+      "loss": 0.334,
+      "step": 2519
+    },
+    {
+      "epoch": 0.24525547445255474,
+      "grad_norm": 1.6342111830003216,
+      "learning_rate": 8.833934477464348e-06,
+      "loss": 0.6127,
+      "step": 2520
+    },
+    {
+      "epoch": 0.24535279805352797,
+      "grad_norm": 1.6465764681762454,
+      "learning_rate": 8.83292255017435e-06,
+      "loss": 0.6432,
+      "step": 2521
+    },
+    {
+      "epoch": 0.24545012165450122,
+      "grad_norm": 1.4262158711234114,
+      "learning_rate": 8.83191024200448e-06,
+      "loss": 0.5224,
+      "step": 2522
+    },
+    {
+      "epoch": 0.24554744525547445,
+      "grad_norm": 1.593193256147642,
+      "learning_rate": 8.830897553055337e-06,
+      "loss": 0.5211,
+      "step": 2523
+    },
+    {
+      "epoch": 0.24564476885644768,
+      "grad_norm": 1.624031218270973,
+      "learning_rate": 8.829884483427547e-06,
+      "loss": 0.5128,
+      "step": 2524
+    },
+    {
+      "epoch": 0.24574209245742093,
+      "grad_norm": 1.991662408778961,
+      "learning_rate": 8.828871033221783e-06,
+      "loss": 0.3025,
+      "step": 2525
+    },
+    {
+      "epoch": 0.24583941605839416,
+      "grad_norm": 1.4390691402915812,
+      "learning_rate": 8.82785720253875e-06,
+      "loss": 0.5088,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2459367396593674,
+      "grad_norm": 1.4179406701872763,
+      "learning_rate": 8.826842991479197e-06,
+      "loss": 0.3887,
+      "step": 2527
+    },
+    {
+      "epoch": 0.24603406326034064,
+      "grad_norm": 1.460230365502962,
+      "learning_rate": 8.825828400143902e-06,
+      "loss": 0.3316,
+      "step": 2528
+    },
+    {
+      "epoch": 0.24613138686131386,
+      "grad_norm": 1.4924241123043909,
+      "learning_rate": 8.824813428633685e-06,
+      "loss": 0.4989,
+      "step": 2529
+    },
+    {
+      "epoch": 0.24622871046228711,
+      "grad_norm": 1.593556186634644,
+      "learning_rate": 8.82379807704941e-06,
+      "loss": 0.525,
+      "step": 2530
+    },
+    {
+      "epoch": 0.24632603406326034,
+      "grad_norm": 1.6809566227650843,
+      "learning_rate": 8.822782345491968e-06,
+      "loss": 0.3421,
+      "step": 2531
+    },
+    {
+      "epoch": 0.24642335766423357,
+      "grad_norm": 1.4773288736144092,
+      "learning_rate": 8.821766234062294e-06,
+      "loss": 0.534,
+      "step": 2532
+    },
+    {
+      "epoch": 0.24652068126520682,
+      "grad_norm": 1.4001059355846526,
+      "learning_rate": 8.820749742861363e-06,
+      "loss": 0.3887,
+      "step": 2533
+    },
+    {
+      "epoch": 0.24661800486618005,
+      "grad_norm": 1.349012582441713,
+      "learning_rate": 8.81973287199018e-06,
+      "loss": 0.2852,
+      "step": 2534
+    },
+    {
+      "epoch": 0.24671532846715327,
+      "grad_norm": 1.651550318908522,
+      "learning_rate": 8.818715621549794e-06,
+      "loss": 0.4967,
+      "step": 2535
+    },
+    {
+      "epoch": 0.24681265206812653,
+      "grad_norm": 1.5932669562049986,
+      "learning_rate": 8.817697991641289e-06,
+      "loss": 0.4173,
+      "step": 2536
+    },
+    {
+      "epoch": 0.24690997566909975,
+      "grad_norm": 1.3550488264007063,
+      "learning_rate": 8.816679982365787e-06,
+      "loss": 0.3404,
+      "step": 2537
+    },
+    {
+      "epoch": 0.247007299270073,
+      "grad_norm": 1.571341106532058,
+      "learning_rate": 8.815661593824451e-06,
+      "loss": 0.5666,
+      "step": 2538
+    },
+    {
+      "epoch": 0.24710462287104623,
+      "grad_norm": 1.5685299297246114,
+      "learning_rate": 8.814642826118477e-06,
+      "loss": 0.4521,
+      "step": 2539
+    },
+    {
+      "epoch": 0.24720194647201946,
+      "grad_norm": 1.5355691524375334,
+      "learning_rate": 8.8136236793491e-06,
+      "loss": 0.3452,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2472992700729927,
+      "grad_norm": 1.4490992247448509,
+      "learning_rate": 8.812604153617594e-06,
+      "loss": 0.3046,
+      "step": 2541
+    },
+    {
+      "epoch": 0.24739659367396594,
+      "grad_norm": 1.5790493967738255,
+      "learning_rate": 8.81158424902527e-06,
+      "loss": 0.5957,
+      "step": 2542
+    },
+    {
+      "epoch": 0.24749391727493916,
+      "grad_norm": 1.8299083651337236,
+      "learning_rate": 8.810563965673478e-06,
+      "loss": 0.529,
+      "step": 2543
+    },
+    {
+      "epoch": 0.24759124087591242,
+      "grad_norm": 1.336357630649535,
+      "learning_rate": 8.8095433036636e-06,
+      "loss": 0.2498,
+      "step": 2544
+    },
+    {
+      "epoch": 0.24768856447688564,
+      "grad_norm": 3.272954864246679,
+      "learning_rate": 8.808522263097065e-06,
+      "loss": 0.3439,
+      "step": 2545
+    },
+    {
+      "epoch": 0.24778588807785887,
+      "grad_norm": 1.5948700054852,
+      "learning_rate": 8.80750084407533e-06,
+      "loss": 0.5754,
+      "step": 2546
+    },
+    {
+      "epoch": 0.24788321167883212,
+      "grad_norm": 1.2457293034288246,
+      "learning_rate": 8.806479046699896e-06,
+      "loss": 0.3355,
+      "step": 2547
+    },
+    {
+      "epoch": 0.24798053527980535,
+      "grad_norm": 1.4118835775208534,
+      "learning_rate": 8.8054568710723e-06,
+      "loss": 0.4843,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2480778588807786,
+      "grad_norm": 2.0167817337794745,
+      "learning_rate": 8.804434317294115e-06,
+      "loss": 0.4781,
+      "step": 2549
+    },
+    {
+      "epoch": 0.24817518248175183,
+      "grad_norm": 1.630746510877536,
+      "learning_rate": 8.803411385466954e-06,
+      "loss": 0.5226,
+      "step": 2550
+    },
+    {
+      "epoch": 0.24827250608272505,
+      "grad_norm": 1.0942598516950242,
+      "learning_rate": 8.802388075692465e-06,
+      "loss": 0.1843,
+      "step": 2551
+    },
+    {
+      "epoch": 0.2483698296836983,
+      "grad_norm": 1.8060042956650721,
+      "learning_rate": 8.801364388072336e-06,
+      "loss": 0.705,
+      "step": 2552
+    },
+    {
+      "epoch": 0.24846715328467153,
+      "grad_norm": 1.632331667833736,
+      "learning_rate": 8.800340322708291e-06,
+      "loss": 0.4964,
+      "step": 2553
+    },
+    {
+      "epoch": 0.24856447688564476,
+      "grad_norm": 1.539098206701319,
+      "learning_rate": 8.799315879702095e-06,
+      "loss": 0.3962,
+      "step": 2554
+    },
+    {
+      "epoch": 0.248661800486618,
+      "grad_norm": 1.2219114137184675,
+      "learning_rate": 8.798291059155543e-06,
+      "loss": 0.2497,
+      "step": 2555
+    },
+    {
+      "epoch": 0.24875912408759124,
+      "grad_norm": 1.4540964796439875,
+      "learning_rate": 8.797265861170471e-06,
+      "loss": 0.5159,
+      "step": 2556
+    },
+    {
+      "epoch": 0.24885644768856446,
+      "grad_norm": 1.554150512584087,
+      "learning_rate": 8.796240285848761e-06,
+      "loss": 0.4412,
+      "step": 2557
+    },
+    {
+      "epoch": 0.24895377128953772,
+      "grad_norm": 1.7004545782091594,
+      "learning_rate": 8.795214333292318e-06,
+      "loss": 0.5179,
+      "step": 2558
+    },
+    {
+      "epoch": 0.24905109489051094,
+      "grad_norm": 1.726524110945535,
+      "learning_rate": 8.794188003603095e-06,
+      "loss": 0.4071,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2491484184914842,
+      "grad_norm": 1.27126477948415,
+      "learning_rate": 8.793161296883077e-06,
+      "loss": 0.2268,
+      "step": 2560
+    },
+    {
+      "epoch": 0.24924574209245742,
+      "grad_norm": 1.9752049062158858,
+      "learning_rate": 8.79213421323429e-06,
+      "loss": 0.3632,
+      "step": 2561
+    },
+    {
+      "epoch": 0.24934306569343065,
+      "grad_norm": 1.0556825817929254,
+      "learning_rate": 8.791106752758796e-06,
+      "loss": 0.3627,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2494403892944039,
+      "grad_norm": 1.6452772754401714,
+      "learning_rate": 8.790078915558693e-06,
+      "loss": 0.6043,
+      "step": 2563
+    },
+    {
+      "epoch": 0.24953771289537713,
+      "grad_norm": 1.278547180886592,
+      "learning_rate": 8.789050701736117e-06,
+      "loss": 0.3768,
+      "step": 2564
+    },
+    {
+      "epoch": 0.24963503649635035,
+      "grad_norm": 1.3443028399521961,
+      "learning_rate": 8.788022111393247e-06,
+      "loss": 0.3856,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2497323600973236,
+      "grad_norm": 1.2774166354695482,
+      "learning_rate": 8.78699314463229e-06,
+      "loss": 0.4391,
+      "step": 2566
+    },
+    {
+      "epoch": 0.24982968369829683,
+      "grad_norm": 1.2231715277397497,
+      "learning_rate": 8.785963801555497e-06,
+      "loss": 0.4128,
+      "step": 2567
+    },
+    {
+      "epoch": 0.24992700729927006,
+      "grad_norm": 1.4012153782510572,
+      "learning_rate": 8.784934082265154e-06,
+      "loss": 0.4683,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2500243309002433,
+      "grad_norm": 1.1954060436870173,
+      "learning_rate": 8.783903986863583e-06,
+      "loss": 0.2786,
+      "step": 2569
+    },
+    {
+      "epoch": 0.25012165450121654,
+      "grad_norm": 1.7116998515615807,
+      "learning_rate": 8.782873515453148e-06,
+      "loss": 0.6004,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2502189781021898,
+      "grad_norm": 1.5712719922889962,
+      "learning_rate": 8.781842668136247e-06,
+      "loss": 0.6172,
+      "step": 2571
+    },
+    {
+      "epoch": 0.250316301703163,
+      "grad_norm": 1.246915874910697,
+      "learning_rate": 8.780811445015316e-06,
+      "loss": 0.4335,
+      "step": 2572
+    },
+    {
+      "epoch": 0.25041362530413624,
+      "grad_norm": 1.341456518636559,
+      "learning_rate": 8.779779846192827e-06,
+      "loss": 0.4187,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2505109489051095,
+      "grad_norm": 1.1323562755710477,
+      "learning_rate": 8.778747871771293e-06,
+      "loss": 0.2832,
+      "step": 2574
+    },
+    {
+      "epoch": 0.25060827250608275,
+      "grad_norm": 1.4401083791532063,
+      "learning_rate": 8.777715521853258e-06,
+      "loss": 0.3779,
+      "step": 2575
+    },
+    {
+      "epoch": 0.25070559610705595,
+      "grad_norm": 1.4784987737181619,
+      "learning_rate": 8.77668279654131e-06,
+      "loss": 0.3129,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2508029197080292,
+      "grad_norm": 1.1394717513462493,
+      "learning_rate": 8.775649695938074e-06,
+      "loss": 0.3162,
+      "step": 2577
+    },
+    {
+      "epoch": 0.25090024330900246,
+      "grad_norm": 1.4625556674372375,
+      "learning_rate": 8.774616220146204e-06,
+      "loss": 0.4605,
+      "step": 2578
+    },
+    {
+      "epoch": 0.25099756690997566,
+      "grad_norm": 3.1521808341091875,
+      "learning_rate": 8.773582369268402e-06,
+      "loss": 0.3485,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2510948905109489,
+      "grad_norm": 1.3578124438111323,
+      "learning_rate": 8.7725481434074e-06,
+      "loss": 0.4693,
+      "step": 2580
+    },
+    {
+      "epoch": 0.25119221411192216,
+      "grad_norm": 1.63411664215404,
+      "learning_rate": 8.771513542665969e-06,
+      "loss": 0.4956,
+      "step": 2581
+    },
+    {
+      "epoch": 0.25128953771289536,
+      "grad_norm": 1.5098765580454843,
+      "learning_rate": 8.77047856714692e-06,
+      "loss": 0.4657,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2513868613138686,
+      "grad_norm": 1.2801786921613054,
+      "learning_rate": 8.7694432169531e-06,
+      "loss": 0.3369,
+      "step": 2583
+    },
+    {
+      "epoch": 0.25148418491484187,
+      "grad_norm": 1.4360422953324754,
+      "learning_rate": 8.768407492187388e-06,
+      "loss": 0.4907,
+      "step": 2584
+    },
+    {
+      "epoch": 0.25158150851581507,
+      "grad_norm": 1.4560406874169747,
+      "learning_rate": 8.767371392952708e-06,
+      "loss": 0.3157,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2516788321167883,
+      "grad_norm": 1.934211832538441,
+      "learning_rate": 8.766334919352018e-06,
+      "loss": 0.7151,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2517761557177616,
+      "grad_norm": 1.6767044903158872,
+      "learning_rate": 8.76529807148831e-06,
+      "loss": 0.331,
+      "step": 2587
+    },
+    {
+      "epoch": 0.25187347931873477,
+      "grad_norm": 1.4698852047894042,
+      "learning_rate": 8.76426084946462e-06,
+      "loss": 0.3951,
+      "step": 2588
+    },
+    {
+      "epoch": 0.251970802919708,
+      "grad_norm": 1.3539539414605721,
+      "learning_rate": 8.763223253384015e-06,
+      "loss": 0.4011,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2520681265206813,
+      "grad_norm": 1.506242240790805,
+      "learning_rate": 8.762185283349603e-06,
+      "loss": 0.5274,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2521654501216545,
+      "grad_norm": 1.3140936667503142,
+      "learning_rate": 8.761146939464527e-06,
+      "loss": 0.3198,
+      "step": 2591
+    },
+    {
+      "epoch": 0.25226277372262773,
+      "grad_norm": 1.1404767919952752,
+      "learning_rate": 8.760108221831967e-06,
+      "loss": 0.4013,
+      "step": 2592
+    },
+    {
+      "epoch": 0.252360097323601,
+      "grad_norm": 1.4693477307137552,
+      "learning_rate": 8.759069130555142e-06,
+      "loss": 0.4783,
+      "step": 2593
+    },
+    {
+      "epoch": 0.2524574209245742,
+      "grad_norm": 1.3352582665983712,
+      "learning_rate": 8.75802966573731e-06,
+      "loss": 0.4617,
+      "step": 2594
+    },
+    {
+      "epoch": 0.25255474452554744,
+      "grad_norm": 1.2824428866870197,
+      "learning_rate": 8.756989827481756e-06,
+      "loss": 0.3352,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2526520681265207,
+      "grad_norm": 1.4774059328965283,
+      "learning_rate": 8.755949615891814e-06,
+      "loss": 0.4635,
+      "step": 2596
+    },
+    {
+      "epoch": 0.25274939172749394,
+      "grad_norm": 1.6875827910282526,
+      "learning_rate": 8.754909031070852e-06,
+      "loss": 0.6222,
+      "step": 2597
+    },
+    {
+      "epoch": 0.25284671532846714,
+      "grad_norm": 1.2063205441417741,
+      "learning_rate": 8.75386807312227e-06,
+      "loss": 0.2455,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2529440389294404,
+      "grad_norm": 1.3021547323360578,
+      "learning_rate": 8.752826742149512e-06,
+      "loss": 0.4329,
+      "step": 2599
+    },
+    {
+      "epoch": 0.25304136253041365,
+      "grad_norm": 1.1835878076183852,
+      "learning_rate": 8.751785038256054e-06,
+      "loss": 0.3662,
+      "step": 2600
+    },
+    {
+      "epoch": 0.25313868613138685,
+      "grad_norm": 1.544717999496196,
+      "learning_rate": 8.750742961545409e-06,
+      "loss": 0.3971,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2532360097323601,
+      "grad_norm": 1.3629505007649398,
+      "learning_rate": 8.749700512121131e-06,
+      "loss": 0.5107,
+      "step": 2602
+    },
+    {
+      "epoch": 0.25333333333333335,
+      "grad_norm": 1.5737001686599814,
+      "learning_rate": 8.74865769008681e-06,
+      "loss": 0.5279,
+      "step": 2603
+    },
+    {
+      "epoch": 0.25343065693430655,
+      "grad_norm": 1.4784815997261378,
+      "learning_rate": 8.747614495546069e-06,
+      "loss": 0.4792,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2535279805352798,
+      "grad_norm": 1.3236076722973804,
+      "learning_rate": 8.74657092860257e-06,
+      "loss": 0.3975,
+      "step": 2605
+    },
+    {
+      "epoch": 0.25362530413625306,
+      "grad_norm": 1.0968595172191475,
+      "learning_rate": 8.745526989360018e-06,
+      "loss": 0.269,
+      "step": 2606
+    },
+    {
+      "epoch": 0.25372262773722626,
+      "grad_norm": 1.562505340567045,
+      "learning_rate": 8.744482677922147e-06,
+      "loss": 0.5157,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2538199513381995,
+      "grad_norm": 1.656826278908397,
+      "learning_rate": 8.743437994392729e-06,
+      "loss": 0.4867,
+      "step": 2608
+    },
+    {
+      "epoch": 0.25391727493917277,
+      "grad_norm": 1.3948672448161548,
+      "learning_rate": 8.742392938875577e-06,
+      "loss": 0.5279,
+      "step": 2609
+    },
+    {
+      "epoch": 0.25401459854014596,
+      "grad_norm": 1.5892338813179163,
+      "learning_rate": 8.741347511474539e-06,
+      "loss": 0.5611,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2541119221411192,
+      "grad_norm": 1.6074249897923194,
+      "learning_rate": 8.740301712293498e-06,
+      "loss": 0.351,
+      "step": 2611
+    },
+    {
+      "epoch": 0.25420924574209247,
+      "grad_norm": 1.6505540033315536,
+      "learning_rate": 8.739255541436379e-06,
+      "loss": 0.5747,
+      "step": 2612
+    },
+    {
+      "epoch": 0.25430656934306567,
+      "grad_norm": 1.3314247428577628,
+      "learning_rate": 8.738208999007137e-06,
+      "loss": 0.3779,
+      "step": 2613
+    },
+    {
+      "epoch": 0.2544038929440389,
+      "grad_norm": 1.2796270745139389,
+      "learning_rate": 8.737162085109768e-06,
+      "loss": 0.3557,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2545012165450122,
+      "grad_norm": 1.602637102567401,
+      "learning_rate": 8.736114799848307e-06,
+      "loss": 0.2882,
+      "step": 2615
+    },
+    {
+      "epoch": 0.2545985401459854,
+      "grad_norm": 1.4207119219562419,
+      "learning_rate": 8.735067143326821e-06,
+      "loss": 0.3881,
+      "step": 2616
+    },
+    {
+      "epoch": 0.25469586374695863,
+      "grad_norm": 1.4305110706379638,
+      "learning_rate": 8.73401911564942e-06,
+      "loss": 0.3486,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2547931873479319,
+      "grad_norm": 1.434428707272536,
+      "learning_rate": 8.732970716920242e-06,
+      "loss": 0.3169,
+      "step": 2618
+    },
+    {
+      "epoch": 0.25489051094890514,
+      "grad_norm": 1.3228470441064362,
+      "learning_rate": 8.73192194724347e-06,
+      "loss": 0.4485,
+      "step": 2619
+    },
+    {
+      "epoch": 0.25498783454987833,
+      "grad_norm": 1.3897030806906485,
+      "learning_rate": 8.730872806723318e-06,
+      "loss": 0.4172,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2550851581508516,
+      "grad_norm": 1.3840681937318722,
+      "learning_rate": 8.729823295464045e-06,
+      "loss": 0.251,
+      "step": 2621
+    },
+    {
+      "epoch": 0.25518248175182484,
+      "grad_norm": 1.775278354364079,
+      "learning_rate": 8.728773413569938e-06,
+      "loss": 0.4811,
+      "step": 2622
+    },
+    {
+      "epoch": 0.25527980535279804,
+      "grad_norm": 1.2701408917829737,
+      "learning_rate": 8.727723161145325e-06,
+      "loss": 0.2827,
+      "step": 2623
+    },
+    {
+      "epoch": 0.2553771289537713,
+      "grad_norm": 1.5528362504659363,
+      "learning_rate": 8.72667253829457e-06,
+      "loss": 0.5084,
+      "step": 2624
+    },
+    {
+      "epoch": 0.25547445255474455,
+      "grad_norm": 1.3793988523162408,
+      "learning_rate": 8.725621545122074e-06,
+      "loss": 0.3979,
+      "step": 2625
+    },
+    {
+      "epoch": 0.25557177615571774,
+      "grad_norm": 1.70282889775673,
+      "learning_rate": 8.724570181732275e-06,
+      "loss": 0.5983,
+      "step": 2626
+    },
+    {
+      "epoch": 0.255669099756691,
+      "grad_norm": 1.28105292316495,
+      "learning_rate": 8.723518448229649e-06,
+      "loss": 0.4756,
+      "step": 2627
+    },
+    {
+      "epoch": 0.25576642335766425,
+      "grad_norm": 1.3826686116158597,
+      "learning_rate": 8.722466344718705e-06,
+      "loss": 0.2978,
+      "step": 2628
+    },
+    {
+      "epoch": 0.25586374695863745,
+      "grad_norm": 1.460242284502631,
+      "learning_rate": 8.721413871303992e-06,
+      "loss": 0.4036,
+      "step": 2629
+    },
+    {
+      "epoch": 0.2559610705596107,
+      "grad_norm": 1.4181157816170762,
+      "learning_rate": 8.720361028090095e-06,
+      "loss": 0.4224,
+      "step": 2630
+    },
+    {
+      "epoch": 0.25605839416058396,
+      "grad_norm": 1.7898330028782403,
+      "learning_rate": 8.719307815181638e-06,
+      "loss": 0.7314,
+      "step": 2631
+    },
+    {
+      "epoch": 0.25615571776155716,
+      "grad_norm": 1.6886124652733636,
+      "learning_rate": 8.718254232683276e-06,
+      "loss": 0.3513,
+      "step": 2632
+    },
+    {
+      "epoch": 0.2562530413625304,
+      "grad_norm": 1.2562027575971086,
+      "learning_rate": 8.717200280699705e-06,
+      "loss": 0.284,
+      "step": 2633
+    },
+    {
+      "epoch": 0.25635036496350366,
+      "grad_norm": 1.4899596514775177,
+      "learning_rate": 8.716145959335658e-06,
+      "loss": 0.2778,
+      "step": 2634
+    },
+    {
+      "epoch": 0.25644768856447686,
+      "grad_norm": 1.1699021581347986,
+      "learning_rate": 8.715091268695903e-06,
+      "loss": 0.3163,
+      "step": 2635
+    },
+    {
+      "epoch": 0.2565450121654501,
+      "grad_norm": 1.020653527182934,
+      "learning_rate": 8.714036208885243e-06,
+      "loss": 0.2191,
+      "step": 2636
+    },
+    {
+      "epoch": 0.25664233576642337,
+      "grad_norm": 1.5373942827305265,
+      "learning_rate": 8.712980780008526e-06,
+      "loss": 0.4183,
+      "step": 2637
+    },
+    {
+      "epoch": 0.25673965936739657,
+      "grad_norm": 1.1268355971062876,
+      "learning_rate": 8.711924982170623e-06,
+      "loss": 0.2851,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2568369829683698,
+      "grad_norm": 1.25228244300652,
+      "learning_rate": 8.710868815476456e-06,
+      "loss": 0.1963,
+      "step": 2639
+    },
+    {
+      "epoch": 0.2569343065693431,
+      "grad_norm": 1.3905442460862172,
+      "learning_rate": 8.709812280030971e-06,
+      "loss": 0.3648,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2570316301703163,
+      "grad_norm": 1.5078176389616522,
+      "learning_rate": 8.708755375939162e-06,
+      "loss": 0.4131,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2571289537712895,
+      "grad_norm": 1.4441200079463874,
+      "learning_rate": 8.70769810330605e-06,
+      "loss": 0.4047,
+      "step": 2642
+    },
+    {
+      "epoch": 0.2572262773722628,
+      "grad_norm": 1.3883503516178042,
+      "learning_rate": 8.7066404622367e-06,
+      "loss": 0.3308,
+      "step": 2643
+    },
+    {
+      "epoch": 0.25732360097323603,
+      "grad_norm": 1.7851696055640995,
+      "learning_rate": 8.705582452836208e-06,
+      "loss": 0.336,
+      "step": 2644
+    },
+    {
+      "epoch": 0.25742092457420923,
+      "grad_norm": 1.309628752016819,
+      "learning_rate": 8.70452407520971e-06,
+      "loss": 0.3462,
+      "step": 2645
+    },
+    {
+      "epoch": 0.2575182481751825,
+      "grad_norm": 1.3618437175125289,
+      "learning_rate": 8.703465329462379e-06,
+      "loss": 0.3047,
+      "step": 2646
+    },
+    {
+      "epoch": 0.25761557177615574,
+      "grad_norm": 1.5821297320572192,
+      "learning_rate": 8.702406215699421e-06,
+      "loss": 0.2318,
+      "step": 2647
+    },
+    {
+      "epoch": 0.25771289537712894,
+      "grad_norm": 1.4729014225467234,
+      "learning_rate": 8.701346734026082e-06,
+      "loss": 0.3147,
+      "step": 2648
+    },
+    {
+      "epoch": 0.2578102189781022,
+      "grad_norm": 1.6287249640343295,
+      "learning_rate": 8.700286884547642e-06,
+      "loss": 0.5808,
+      "step": 2649
+    },
+    {
+      "epoch": 0.25790754257907544,
+      "grad_norm": 1.2824109098190504,
+      "learning_rate": 8.69922666736942e-06,
+      "loss": 0.3836,
+      "step": 2650
+    },
+    {
+      "epoch": 0.25800486618004864,
+      "grad_norm": 1.5096397594183033,
+      "learning_rate": 8.69816608259677e-06,
+      "loss": 0.3804,
+      "step": 2651
+    },
+    {
+      "epoch": 0.2581021897810219,
+      "grad_norm": 1.7247008216261863,
+      "learning_rate": 8.697105130335084e-06,
+      "loss": 0.3378,
+      "step": 2652
+    },
+    {
+      "epoch": 0.25819951338199515,
+      "grad_norm": 1.5872130127065738,
+      "learning_rate": 8.69604381068979e-06,
+      "loss": 0.4369,
+      "step": 2653
+    },
+    {
+      "epoch": 0.25829683698296835,
+      "grad_norm": 1.5909295650502344,
+      "learning_rate": 8.694982123766348e-06,
+      "loss": 0.3554,
+      "step": 2654
+    },
+    {
+      "epoch": 0.2583941605839416,
+      "grad_norm": 1.7135035115393307,
+      "learning_rate": 8.693920069670265e-06,
+      "loss": 0.4869,
+      "step": 2655
+    },
+    {
+      "epoch": 0.25849148418491485,
+      "grad_norm": 1.3366492087792976,
+      "learning_rate": 8.692857648507071e-06,
+      "loss": 0.3102,
+      "step": 2656
+    },
+    {
+      "epoch": 0.25858880778588805,
+      "grad_norm": 1.2478048122674565,
+      "learning_rate": 8.691794860382345e-06,
+      "loss": 0.3722,
+      "step": 2657
+    },
+    {
+      "epoch": 0.2586861313868613,
+      "grad_norm": 1.5080776475601503,
+      "learning_rate": 8.690731705401694e-06,
+      "loss": 0.316,
+      "step": 2658
+    },
+    {
+      "epoch": 0.25878345498783456,
+      "grad_norm": 1.443811575497146,
+      "learning_rate": 8.689668183670763e-06,
+      "loss": 0.2875,
+      "step": 2659
+    },
+    {
+      "epoch": 0.25888077858880776,
+      "grad_norm": 1.7036441396737687,
+      "learning_rate": 8.688604295295238e-06,
+      "loss": 0.4025,
+      "step": 2660
+    },
+    {
+      "epoch": 0.258978102189781,
+      "grad_norm": 1.4234806259439374,
+      "learning_rate": 8.687540040380838e-06,
+      "loss": 0.4452,
+      "step": 2661
+    },
+    {
+      "epoch": 0.25907542579075427,
+      "grad_norm": 1.2741393980838642,
+      "learning_rate": 8.686475419033315e-06,
+      "loss": 0.2237,
+      "step": 2662
+    },
+    {
+      "epoch": 0.2591727493917275,
+      "grad_norm": 1.1826384563722763,
+      "learning_rate": 8.685410431358464e-06,
+      "loss": 0.3398,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2592700729927007,
+      "grad_norm": 1.5757741509023746,
+      "learning_rate": 8.684345077462117e-06,
+      "loss": 0.3846,
+      "step": 2664
+    },
+    {
+      "epoch": 0.25936739659367397,
+      "grad_norm": 1.475707275733763,
+      "learning_rate": 8.683279357450131e-06,
+      "loss": 0.2804,
+      "step": 2665
+    },
+    {
+      "epoch": 0.2594647201946472,
+      "grad_norm": 1.4241797244636094,
+      "learning_rate": 8.682213271428415e-06,
+      "loss": 0.2553,
+      "step": 2666
+    },
+    {
+      "epoch": 0.2595620437956204,
+      "grad_norm": 1.1548194283365685,
+      "learning_rate": 8.6811468195029e-06,
+      "loss": 0.3118,
+      "step": 2667
+    },
+    {
+      "epoch": 0.2596593673965937,
+      "grad_norm": 1.5918458521510486,
+      "learning_rate": 8.680080001779564e-06,
+      "loss": 0.4525,
+      "step": 2668
+    },
+    {
+      "epoch": 0.25975669099756693,
+      "grad_norm": 1.5508802560099362,
+      "learning_rate": 8.679012818364416e-06,
+      "loss": 0.4163,
+      "step": 2669
+    },
+    {
+      "epoch": 0.25985401459854013,
+      "grad_norm": 2.4434630008376232,
+      "learning_rate": 8.677945269363504e-06,
+      "loss": 0.4372,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2599513381995134,
+      "grad_norm": 1.5324792404386718,
+      "learning_rate": 8.676877354882907e-06,
+      "loss": 0.3514,
+      "step": 2671
+    },
+    {
+      "epoch": 0.26004866180048664,
+      "grad_norm": 2.0012246197360493,
+      "learning_rate": 8.67580907502875e-06,
+      "loss": 0.5067,
+      "step": 2672
+    },
+    {
+      "epoch": 0.26014598540145983,
+      "grad_norm": 1.5232176793280576,
+      "learning_rate": 8.674740429907186e-06,
+      "loss": 0.4174,
+      "step": 2673
+    },
+    {
+      "epoch": 0.2602433090024331,
+      "grad_norm": 1.3322865976928646,
+      "learning_rate": 8.673671419624405e-06,
+      "loss": 0.4095,
+      "step": 2674
+    },
+    {
+      "epoch": 0.26034063260340634,
+      "grad_norm": 1.514406481268828,
+      "learning_rate": 8.672602044286638e-06,
+      "loss": 0.5915,
+      "step": 2675
+    },
+    {
+      "epoch": 0.26043795620437954,
+      "grad_norm": 1.528467413797325,
+      "learning_rate": 8.67153230400015e-06,
+      "loss": 0.4018,
+      "step": 2676
+    },
+    {
+      "epoch": 0.2605352798053528,
+      "grad_norm": 1.4367698805538582,
+      "learning_rate": 8.670462198871237e-06,
+      "loss": 0.4115,
+      "step": 2677
+    },
+    {
+      "epoch": 0.26063260340632605,
+      "grad_norm": 1.6984444092554742,
+      "learning_rate": 8.66939172900624e-06,
+      "loss": 0.59,
+      "step": 2678
+    },
+    {
+      "epoch": 0.26072992700729924,
+      "grad_norm": 1.4698751482200727,
+      "learning_rate": 8.668320894511534e-06,
+      "loss": 0.4144,
+      "step": 2679
+    },
+    {
+      "epoch": 0.2608272506082725,
+      "grad_norm": 1.5003641004534345,
+      "learning_rate": 8.667249695493525e-06,
+      "loss": 0.4294,
+      "step": 2680
+    },
+    {
+      "epoch": 0.26092457420924575,
+      "grad_norm": 1.3123452231563197,
+      "learning_rate": 8.666178132058659e-06,
+      "loss": 0.3408,
+      "step": 2681
+    },
+    {
+      "epoch": 0.261021897810219,
+      "grad_norm": 1.5184535738040659,
+      "learning_rate": 8.665106204313418e-06,
+      "loss": 0.3662,
+      "step": 2682
+    },
+    {
+      "epoch": 0.2611192214111922,
+      "grad_norm": 1.0623024588559944,
+      "learning_rate": 8.664033912364321e-06,
+      "loss": 0.2953,
+      "step": 2683
+    },
+    {
+      "epoch": 0.26121654501216546,
+      "grad_norm": 1.4112725317400583,
+      "learning_rate": 8.662961256317923e-06,
+      "loss": 0.3825,
+      "step": 2684
+    },
+    {
+      "epoch": 0.2613138686131387,
+      "grad_norm": 2.2729536767377065,
+      "learning_rate": 8.661888236280813e-06,
+      "loss": 0.5791,
+      "step": 2685
+    },
+    {
+      "epoch": 0.2614111922141119,
+      "grad_norm": 2.2747614305768504,
+      "learning_rate": 8.660814852359617e-06,
+      "loss": 0.4859,
+      "step": 2686
+    },
+    {
+      "epoch": 0.26150851581508516,
+      "grad_norm": 1.6069562939941755,
+      "learning_rate": 8.659741104661002e-06,
+      "loss": 0.5254,
+      "step": 2687
+    },
+    {
+      "epoch": 0.2616058394160584,
+      "grad_norm": 1.3624858995460438,
+      "learning_rate": 8.658666993291662e-06,
+      "loss": 0.3904,
+      "step": 2688
+    },
+    {
+      "epoch": 0.2617031630170316,
+      "grad_norm": 1.2954398797770197,
+      "learning_rate": 8.657592518358332e-06,
+      "loss": 0.3789,
+      "step": 2689
+    },
+    {
+      "epoch": 0.26180048661800487,
+      "grad_norm": 1.4158991903907718,
+      "learning_rate": 8.656517679967788e-06,
+      "loss": 0.3732,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2618978102189781,
+      "grad_norm": 1.3754641009755615,
+      "learning_rate": 8.655442478226835e-06,
+      "loss": 0.3035,
+      "step": 2691
+    },
+    {
+      "epoch": 0.2619951338199513,
+      "grad_norm": 1.3522608722257456,
+      "learning_rate": 8.654366913242316e-06,
+      "loss": 0.347,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2620924574209246,
+      "grad_norm": 1.2764013704656585,
+      "learning_rate": 8.65329098512111e-06,
+      "loss": 0.4207,
+      "step": 2693
+    },
+    {
+      "epoch": 0.2621897810218978,
+      "grad_norm": 1.4009476621873176,
+      "learning_rate": 8.652214693970133e-06,
+      "loss": 0.4628,
+      "step": 2694
+    },
+    {
+      "epoch": 0.262287104622871,
+      "grad_norm": 1.3860597575903169,
+      "learning_rate": 8.65113803989634e-06,
+      "loss": 0.3844,
+      "step": 2695
+    },
+    {
+      "epoch": 0.2623844282238443,
+      "grad_norm": 1.5636622874346966,
+      "learning_rate": 8.650061023006711e-06,
+      "loss": 0.6239,
+      "step": 2696
+    },
+    {
+      "epoch": 0.26248175182481753,
+      "grad_norm": 1.3677003606993399,
+      "learning_rate": 8.648983643408276e-06,
+      "loss": 0.4319,
+      "step": 2697
+    },
+    {
+      "epoch": 0.26257907542579073,
+      "grad_norm": 1.4720449620822884,
+      "learning_rate": 8.647905901208096e-06,
+      "loss": 0.4824,
+      "step": 2698
+    },
+    {
+      "epoch": 0.262676399026764,
+      "grad_norm": 1.4180687903221385,
+      "learning_rate": 8.646827796513262e-06,
+      "loss": 0.539,
+      "step": 2699
+    },
+    {
+      "epoch": 0.26277372262773724,
+      "grad_norm": 1.3679667840460958,
+      "learning_rate": 8.64574932943091e-06,
+      "loss": 0.4588,
+      "step": 2700
+    },
+    {
+      "epoch": 0.26287104622871044,
+      "grad_norm": 1.125542933529368,
+      "learning_rate": 8.644670500068205e-06,
+      "loss": 0.3441,
+      "step": 2701
+    },
+    {
+      "epoch": 0.2629683698296837,
+      "grad_norm": 1.5641789380613262,
+      "learning_rate": 8.643591308532353e-06,
+      "loss": 0.4998,
+      "step": 2702
+    },
+    {
+      "epoch": 0.26306569343065694,
+      "grad_norm": 1.3425870342919086,
+      "learning_rate": 8.642511754930592e-06,
+      "loss": 0.4678,
+      "step": 2703
+    },
+    {
+      "epoch": 0.2631630170316302,
+      "grad_norm": 1.3010588112101855,
+      "learning_rate": 8.641431839370199e-06,
+      "loss": 0.4005,
+      "step": 2704
+    },
+    {
+      "epoch": 0.2632603406326034,
+      "grad_norm": 1.0067860306988832,
+      "learning_rate": 8.640351561958487e-06,
+      "loss": 0.2243,
+      "step": 2705
+    },
+    {
+      "epoch": 0.26335766423357665,
+      "grad_norm": 1.4713856201410829,
+      "learning_rate": 8.639270922802802e-06,
+      "loss": 0.4325,
+      "step": 2706
+    },
+    {
+      "epoch": 0.2634549878345499,
+      "grad_norm": 1.55962351192921,
+      "learning_rate": 8.63818992201053e-06,
+      "loss": 0.5307,
+      "step": 2707
+    },
+    {
+      "epoch": 0.2635523114355231,
+      "grad_norm": 1.4073629002175063,
+      "learning_rate": 8.637108559689088e-06,
+      "loss": 0.3329,
+      "step": 2708
+    },
+    {
+      "epoch": 0.26364963503649635,
+      "grad_norm": 1.2827086170801953,
+      "learning_rate": 8.636026835945933e-06,
+      "loss": 0.3095,
+      "step": 2709
+    },
+    {
+      "epoch": 0.2637469586374696,
+      "grad_norm": 1.4100209855486194,
+      "learning_rate": 8.634944750888556e-06,
+      "loss": 0.3033,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2638442822384428,
+      "grad_norm": 1.343279822840104,
+      "learning_rate": 8.633862304624484e-06,
+      "loss": 0.402,
+      "step": 2711
+    },
+    {
+      "epoch": 0.26394160583941606,
+      "grad_norm": 1.4374516520455163,
+      "learning_rate": 8.632779497261284e-06,
+      "loss": 0.4574,
+      "step": 2712
+    },
+    {
+      "epoch": 0.2640389294403893,
+      "grad_norm": 1.1554648336740065,
+      "learning_rate": 8.63169632890655e-06,
+      "loss": 0.3091,
+      "step": 2713
+    },
+    {
+      "epoch": 0.2641362530413625,
+      "grad_norm": 1.5304191047752203,
+      "learning_rate": 8.630612799667923e-06,
+      "loss": 0.5392,
+      "step": 2714
+    },
+    {
+      "epoch": 0.26423357664233577,
+      "grad_norm": 1.9364214941018973,
+      "learning_rate": 8.629528909653067e-06,
+      "loss": 0.4705,
+      "step": 2715
+    },
+    {
+      "epoch": 0.264330900243309,
+      "grad_norm": 1.5176007479008755,
+      "learning_rate": 8.628444658969694e-06,
+      "loss": 0.3969,
+      "step": 2716
+    },
+    {
+      "epoch": 0.2644282238442822,
+      "grad_norm": 1.3882529784475808,
+      "learning_rate": 8.627360047725543e-06,
+      "loss": 0.4672,
+      "step": 2717
+    },
+    {
+      "epoch": 0.26452554744525547,
+      "grad_norm": 1.0419873824719341,
+      "learning_rate": 8.626275076028397e-06,
+      "loss": 0.2247,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2646228710462287,
+      "grad_norm": 1.4147177174052021,
+      "learning_rate": 8.625189743986068e-06,
+      "loss": 0.3922,
+      "step": 2719
+    },
+    {
+      "epoch": 0.2647201946472019,
+      "grad_norm": 1.3513629744004096,
+      "learning_rate": 8.624104051706405e-06,
+      "loss": 0.415,
+      "step": 2720
+    },
+    {
+      "epoch": 0.2648175182481752,
+      "grad_norm": 1.3701041364422066,
+      "learning_rate": 8.623017999297294e-06,
+      "loss": 0.4329,
+      "step": 2721
+    },
+    {
+      "epoch": 0.26491484184914843,
+      "grad_norm": 1.5102917148163044,
+      "learning_rate": 8.621931586866658e-06,
+      "loss": 0.4104,
+      "step": 2722
+    },
+    {
+      "epoch": 0.26501216545012163,
+      "grad_norm": 1.4836677874290423,
+      "learning_rate": 8.620844814522455e-06,
+      "loss": 0.5131,
+      "step": 2723
+    },
+    {
+      "epoch": 0.2651094890510949,
+      "grad_norm": 1.2607364196409017,
+      "learning_rate": 8.619757682372675e-06,
+      "loss": 0.3856,
+      "step": 2724
+    },
+    {
+      "epoch": 0.26520681265206814,
+      "grad_norm": 1.4082529003642341,
+      "learning_rate": 8.61867019052535e-06,
+      "loss": 0.4719,
+      "step": 2725
+    },
+    {
+      "epoch": 0.2653041362530414,
+      "grad_norm": 1.4276001080419702,
+      "learning_rate": 8.617582339088545e-06,
+      "loss": 0.2825,
+      "step": 2726
+    },
+    {
+      "epoch": 0.2654014598540146,
+      "grad_norm": 1.4331001450603844,
+      "learning_rate": 8.61649412817036e-06,
+      "loss": 0.5104,
+      "step": 2727
+    },
+    {
+      "epoch": 0.26549878345498784,
+      "grad_norm": 1.358868383954866,
+      "learning_rate": 8.615405557878929e-06,
+      "loss": 0.4359,
+      "step": 2728
+    },
+    {
+      "epoch": 0.2655961070559611,
+      "grad_norm": 1.678463370024911,
+      "learning_rate": 8.614316628322427e-06,
+      "loss": 0.4658,
+      "step": 2729
+    },
+    {
+      "epoch": 0.2656934306569343,
+      "grad_norm": 1.2268291596580612,
+      "learning_rate": 8.61322733960906e-06,
+      "loss": 0.2337,
+      "step": 2730
+    },
+    {
+      "epoch": 0.26579075425790755,
+      "grad_norm": 0.9437944818586388,
+      "learning_rate": 8.61213769184707e-06,
+      "loss": 0.2525,
+      "step": 2731
+    },
+    {
+      "epoch": 0.2658880778588808,
+      "grad_norm": 1.2480121542051432,
+      "learning_rate": 8.611047685144737e-06,
+      "loss": 0.2656,
+      "step": 2732
+    },
+    {
+      "epoch": 0.265985401459854,
+      "grad_norm": 1.5255853623894704,
+      "learning_rate": 8.609957319610377e-06,
+      "loss": 0.5071,
+      "step": 2733
+    },
+    {
+      "epoch": 0.26608272506082725,
+      "grad_norm": 1.5847632660353408,
+      "learning_rate": 8.60886659535234e-06,
+      "loss": 0.4018,
+      "step": 2734
+    },
+    {
+      "epoch": 0.2661800486618005,
+      "grad_norm": 1.3469310633769445,
+      "learning_rate": 8.60777551247901e-06,
+      "loss": 0.451,
+      "step": 2735
+    },
+    {
+      "epoch": 0.2662773722627737,
+      "grad_norm": 1.3995570810499534,
+      "learning_rate": 8.60668407109881e-06,
+      "loss": 0.4991,
+      "step": 2736
+    },
+    {
+      "epoch": 0.26637469586374696,
+      "grad_norm": 1.5198269828404072,
+      "learning_rate": 8.605592271320199e-06,
+      "loss": 0.4266,
+      "step": 2737
+    },
+    {
+      "epoch": 0.2664720194647202,
+      "grad_norm": 1.3040716122405567,
+      "learning_rate": 8.604500113251666e-06,
+      "loss": 0.3465,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2665693430656934,
+      "grad_norm": 1.3643506509353014,
+      "learning_rate": 8.60340759700174e-06,
+      "loss": 0.4355,
+      "step": 2739
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.026074804296968,
+      "learning_rate": 8.602314722678989e-06,
+      "loss": 0.2507,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2667639902676399,
+      "grad_norm": 1.3894972782664292,
+      "learning_rate": 8.601221490392009e-06,
+      "loss": 0.3981,
+      "step": 2741
+    },
+    {
+      "epoch": 0.2668613138686131,
+      "grad_norm": 1.3071238902768438,
+      "learning_rate": 8.600127900249435e-06,
+      "loss": 0.5138,
+      "step": 2742
+    },
+    {
+      "epoch": 0.26695863746958637,
+      "grad_norm": 1.61583752885221,
+      "learning_rate": 8.59903395235994e-06,
+      "loss": 0.5072,
+      "step": 2743
+    },
+    {
+      "epoch": 0.2670559610705596,
+      "grad_norm": 1.3679578518174673,
+      "learning_rate": 8.597939646832227e-06,
+      "loss": 0.3754,
+      "step": 2744
+    },
+    {
+      "epoch": 0.2671532846715328,
+      "grad_norm": 1.0943121419181938,
+      "learning_rate": 8.596844983775042e-06,
+      "loss": 0.2457,
+      "step": 2745
+    },
+    {
+      "epoch": 0.2672506082725061,
+      "grad_norm": 1.176479145152164,
+      "learning_rate": 8.59574996329716e-06,
+      "loss": 0.3687,
+      "step": 2746
+    },
+    {
+      "epoch": 0.2673479318734793,
+      "grad_norm": 1.2666642902167933,
+      "learning_rate": 8.594654585507393e-06,
+      "loss": 0.2664,
+      "step": 2747
+    },
+    {
+      "epoch": 0.2674452554744526,
+      "grad_norm": 1.3951377938692817,
+      "learning_rate": 8.59355885051459e-06,
+      "loss": 0.4035,
+      "step": 2748
+    },
+    {
+      "epoch": 0.2675425790754258,
+      "grad_norm": 1.2722832533001889,
+      "learning_rate": 8.592462758427635e-06,
+      "loss": 0.4643,
+      "step": 2749
+    },
+    {
+      "epoch": 0.26763990267639903,
+      "grad_norm": 1.2157588835981379,
+      "learning_rate": 8.59136630935545e-06,
+      "loss": 0.3612,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2677372262773723,
+      "grad_norm": 1.0785566378114326,
+      "learning_rate": 8.590269503406986e-06,
+      "loss": 0.3403,
+      "step": 2751
+    },
+    {
+      "epoch": 0.2678345498783455,
+      "grad_norm": 1.2447292785758555,
+      "learning_rate": 8.589172340691235e-06,
+      "loss": 0.3873,
+      "step": 2752
+    },
+    {
+      "epoch": 0.26793187347931874,
+      "grad_norm": 1.166378916722292,
+      "learning_rate": 8.588074821317222e-06,
+      "loss": 0.3264,
+      "step": 2753
+    },
+    {
+      "epoch": 0.268029197080292,
+      "grad_norm": 1.2197572995933224,
+      "learning_rate": 8.586976945394008e-06,
+      "loss": 0.3793,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2681265206812652,
+      "grad_norm": 1.6234832434134598,
+      "learning_rate": 8.58587871303069e-06,
+      "loss": 0.5521,
+      "step": 2755
+    },
+    {
+      "epoch": 0.26822384428223844,
+      "grad_norm": 1.4760533014923396,
+      "learning_rate": 8.584780124336403e-06,
+      "loss": 0.5024,
+      "step": 2756
+    },
+    {
+      "epoch": 0.2683211678832117,
+      "grad_norm": 1.4156240197993037,
+      "learning_rate": 8.58368117942031e-06,
+      "loss": 0.2848,
+      "step": 2757
+    },
+    {
+      "epoch": 0.2684184914841849,
+      "grad_norm": 1.9092848960981135,
+      "learning_rate": 8.582581878391614e-06,
+      "loss": 0.4053,
+      "step": 2758
+    },
+    {
+      "epoch": 0.26851581508515815,
+      "grad_norm": 1.2158050168465575,
+      "learning_rate": 8.581482221359557e-06,
+      "loss": 0.2709,
+      "step": 2759
+    },
+    {
+      "epoch": 0.2686131386861314,
+      "grad_norm": 1.5515245630825936,
+      "learning_rate": 8.580382208433408e-06,
+      "loss": 0.4549,
+      "step": 2760
+    },
+    {
+      "epoch": 0.2687104622871046,
+      "grad_norm": 1.6603384837941395,
+      "learning_rate": 8.57928183972248e-06,
+      "loss": 0.3316,
+      "step": 2761
+    },
+    {
+      "epoch": 0.26880778588807785,
+      "grad_norm": 1.5595744401068579,
+      "learning_rate": 8.578181115336114e-06,
+      "loss": 0.5733,
+      "step": 2762
+    },
+    {
+      "epoch": 0.2689051094890511,
+      "grad_norm": 1.3547786308004384,
+      "learning_rate": 8.577080035383693e-06,
+      "loss": 0.5295,
+      "step": 2763
+    },
+    {
+      "epoch": 0.2690024330900243,
+      "grad_norm": 1.2889595684224195,
+      "learning_rate": 8.57597859997463e-06,
+      "loss": 0.3876,
+      "step": 2764
+    },
+    {
+      "epoch": 0.26909975669099756,
+      "grad_norm": 1.5401948742368967,
+      "learning_rate": 8.574876809218375e-06,
+      "loss": 0.4847,
+      "step": 2765
+    },
+    {
+      "epoch": 0.2691970802919708,
+      "grad_norm": 1.5886773556984544,
+      "learning_rate": 8.573774663224414e-06,
+      "loss": 0.4746,
+      "step": 2766
+    },
+    {
+      "epoch": 0.269294403892944,
+      "grad_norm": 1.2747463684628804,
+      "learning_rate": 8.572672162102269e-06,
+      "loss": 0.2568,
+      "step": 2767
+    },
+    {
+      "epoch": 0.26939172749391727,
+      "grad_norm": 1.1674673988315882,
+      "learning_rate": 8.571569305961495e-06,
+      "loss": 0.4329,
+      "step": 2768
+    },
+    {
+      "epoch": 0.2694890510948905,
+      "grad_norm": 1.6882113617461265,
+      "learning_rate": 8.570466094911684e-06,
+      "loss": 0.6891,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2695863746958638,
+      "grad_norm": 1.6660737969996857,
+      "learning_rate": 8.569362529062461e-06,
+      "loss": 0.5887,
+      "step": 2770
+    },
+    {
+      "epoch": 0.26968369829683697,
+      "grad_norm": 1.1653044559020052,
+      "learning_rate": 8.568258608523491e-06,
+      "loss": 0.2452,
+      "step": 2771
+    },
+    {
+      "epoch": 0.2697810218978102,
+      "grad_norm": 1.5681206888540218,
+      "learning_rate": 8.567154333404471e-06,
+      "loss": 0.4952,
+      "step": 2772
+    },
+    {
+      "epoch": 0.2698783454987835,
+      "grad_norm": 1.3994591247160806,
+      "learning_rate": 8.56604970381513e-06,
+      "loss": 0.2848,
+      "step": 2773
+    },
+    {
+      "epoch": 0.2699756690997567,
+      "grad_norm": 1.300192393224716,
+      "learning_rate": 8.564944719865238e-06,
+      "loss": 0.3924,
+      "step": 2774
+    },
+    {
+      "epoch": 0.27007299270072993,
+      "grad_norm": 1.4412015443912716,
+      "learning_rate": 8.5638393816646e-06,
+      "loss": 0.4531,
+      "step": 2775
+    },
+    {
+      "epoch": 0.2701703163017032,
+      "grad_norm": 1.4360872043281558,
+      "learning_rate": 8.56273368932305e-06,
+      "loss": 0.4571,
+      "step": 2776
+    },
+    {
+      "epoch": 0.2702676399026764,
+      "grad_norm": 1.5811581309774965,
+      "learning_rate": 8.561627642950465e-06,
+      "loss": 0.4638,
+      "step": 2777
+    },
+    {
+      "epoch": 0.27036496350364964,
+      "grad_norm": 1.7924696283680308,
+      "learning_rate": 8.560521242656751e-06,
+      "loss": 0.2922,
+      "step": 2778
+    },
+    {
+      "epoch": 0.2704622871046229,
+      "grad_norm": 1.7929283253885162,
+      "learning_rate": 8.559414488551854e-06,
+      "loss": 0.6197,
+      "step": 2779
+    },
+    {
+      "epoch": 0.2705596107055961,
+      "grad_norm": 1.5593955671219286,
+      "learning_rate": 8.558307380745751e-06,
+      "loss": 0.5448,
+      "step": 2780
+    },
+    {
+      "epoch": 0.27065693430656934,
+      "grad_norm": 1.3760682204767343,
+      "learning_rate": 8.557199919348455e-06,
+      "loss": 0.4434,
+      "step": 2781
+    },
+    {
+      "epoch": 0.2707542579075426,
+      "grad_norm": 3.203989647256839,
+      "learning_rate": 8.556092104470019e-06,
+      "loss": 0.4323,
+      "step": 2782
+    },
+    {
+      "epoch": 0.2708515815085158,
+      "grad_norm": 1.3460764595466628,
+      "learning_rate": 8.554983936220525e-06,
+      "loss": 0.3367,
+      "step": 2783
+    },
+    {
+      "epoch": 0.27094890510948905,
+      "grad_norm": 1.6160732245190643,
+      "learning_rate": 8.553875414710088e-06,
+      "loss": 0.5301,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2710462287104623,
+      "grad_norm": 1.5749454761331767,
+      "learning_rate": 8.552766540048872e-06,
+      "loss": 0.3741,
+      "step": 2785
+    },
+    {
+      "epoch": 0.2711435523114355,
+      "grad_norm": 1.150423059184381,
+      "learning_rate": 8.551657312347057e-06,
+      "loss": 0.2796,
+      "step": 2786
+    },
+    {
+      "epoch": 0.27124087591240875,
+      "grad_norm": 1.4217054664233575,
+      "learning_rate": 8.550547731714874e-06,
+      "loss": 0.4543,
+      "step": 2787
+    },
+    {
+      "epoch": 0.271338199513382,
+      "grad_norm": 1.470206005686861,
+      "learning_rate": 8.54943779826258e-06,
+      "loss": 0.438,
+      "step": 2788
+    },
+    {
+      "epoch": 0.2714355231143552,
+      "grad_norm": 1.5766219733733982,
+      "learning_rate": 8.54832751210047e-06,
+      "loss": 0.4966,
+      "step": 2789
+    },
+    {
+      "epoch": 0.27153284671532846,
+      "grad_norm": 1.2135102045567707,
+      "learning_rate": 8.547216873338876e-06,
+      "loss": 0.358,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2716301703163017,
+      "grad_norm": 1.4595225616938101,
+      "learning_rate": 8.546105882088158e-06,
+      "loss": 0.2225,
+      "step": 2791
+    },
+    {
+      "epoch": 0.27172749391727496,
+      "grad_norm": 1.3363330099445299,
+      "learning_rate": 8.54499453845872e-06,
+      "loss": 0.3914,
+      "step": 2792
+    },
+    {
+      "epoch": 0.27182481751824816,
+      "grad_norm": 1.3646141902938869,
+      "learning_rate": 8.543882842560997e-06,
+      "loss": 0.4558,
+      "step": 2793
+    },
+    {
+      "epoch": 0.2719221411192214,
+      "grad_norm": 1.3464180828493995,
+      "learning_rate": 8.542770794505456e-06,
+      "loss": 0.4786,
+      "step": 2794
+    },
+    {
+      "epoch": 0.27201946472019467,
+      "grad_norm": 1.044551377255888,
+      "learning_rate": 8.541658394402606e-06,
+      "loss": 0.303,
+      "step": 2795
+    },
+    {
+      "epoch": 0.27211678832116787,
+      "grad_norm": 1.6706499263846184,
+      "learning_rate": 8.540545642362982e-06,
+      "loss": 0.4033,
+      "step": 2796
+    },
+    {
+      "epoch": 0.2722141119221411,
+      "grad_norm": 1.3164784669169094,
+      "learning_rate": 8.539432538497162e-06,
+      "loss": 0.4343,
+      "step": 2797
+    },
+    {
+      "epoch": 0.2723114355231144,
+      "grad_norm": 1.6044535524867656,
+      "learning_rate": 8.538319082915757e-06,
+      "loss": 0.3641,
+      "step": 2798
+    },
+    {
+      "epoch": 0.2724087591240876,
+      "grad_norm": 1.9897822202433566,
+      "learning_rate": 8.537205275729406e-06,
+      "loss": 0.48,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2725060827250608,
+      "grad_norm": 1.4110579632506512,
+      "learning_rate": 8.536091117048794e-06,
+      "loss": 0.4798,
+      "step": 2800
+    },
+    {
+      "epoch": 0.2726034063260341,
+      "grad_norm": 1.4415607317920478,
+      "learning_rate": 8.534976606984636e-06,
+      "loss": 0.343,
+      "step": 2801
+    },
+    {
+      "epoch": 0.2727007299270073,
+      "grad_norm": 1.6363482727427716,
+      "learning_rate": 8.53386174564768e-06,
+      "loss": 0.6087,
+      "step": 2802
+    },
+    {
+      "epoch": 0.27279805352798053,
+      "grad_norm": 1.1272383780084416,
+      "learning_rate": 8.532746533148708e-06,
+      "loss": 0.2444,
+      "step": 2803
+    },
+    {
+      "epoch": 0.2728953771289538,
+      "grad_norm": 1.712140222332907,
+      "learning_rate": 8.531630969598544e-06,
+      "loss": 0.6702,
+      "step": 2804
+    },
+    {
+      "epoch": 0.272992700729927,
+      "grad_norm": 1.474485197586056,
+      "learning_rate": 8.530515055108038e-06,
+      "loss": 0.3876,
+      "step": 2805
+    },
+    {
+      "epoch": 0.27309002433090024,
+      "grad_norm": 1.2926370708159094,
+      "learning_rate": 8.529398789788082e-06,
+      "loss": 0.3239,
+      "step": 2806
+    },
+    {
+      "epoch": 0.2731873479318735,
+      "grad_norm": 1.1171205940753008,
+      "learning_rate": 8.528282173749599e-06,
+      "loss": 0.3135,
+      "step": 2807
+    },
+    {
+      "epoch": 0.2732846715328467,
+      "grad_norm": 1.3561762741371761,
+      "learning_rate": 8.527165207103546e-06,
+      "loss": 0.4686,
+      "step": 2808
+    },
+    {
+      "epoch": 0.27338199513381994,
+      "grad_norm": 1.3082129080843141,
+      "learning_rate": 8.52604788996092e-06,
+      "loss": 0.4274,
+      "step": 2809
+    },
+    {
+      "epoch": 0.2734793187347932,
+      "grad_norm": 1.2958697823961909,
+      "learning_rate": 8.524930222432748e-06,
+      "loss": 0.4334,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2735766423357664,
+      "grad_norm": 1.4541266485936315,
+      "learning_rate": 8.523812204630093e-06,
+      "loss": 0.5685,
+      "step": 2811
+    },
+    {
+      "epoch": 0.27367396593673965,
+      "grad_norm": 1.3303596097899522,
+      "learning_rate": 8.522693836664052e-06,
+      "loss": 0.4305,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2737712895377129,
+      "grad_norm": 1.220005269273729,
+      "learning_rate": 8.521575118645761e-06,
+      "loss": 0.4281,
+      "step": 2813
+    },
+    {
+      "epoch": 0.27386861313868616,
+      "grad_norm": 1.0981673276035366,
+      "learning_rate": 8.520456050686384e-06,
+      "loss": 0.3641,
+      "step": 2814
+    },
+    {
+      "epoch": 0.27396593673965935,
+      "grad_norm": 1.4310281439998578,
+      "learning_rate": 8.519336632897128e-06,
+      "loss": 0.557,
+      "step": 2815
+    },
+    {
+      "epoch": 0.2740632603406326,
+      "grad_norm": 1.345841620727785,
+      "learning_rate": 8.518216865389227e-06,
+      "loss": 0.3991,
+      "step": 2816
+    },
+    {
+      "epoch": 0.27416058394160586,
+      "grad_norm": 1.6650753610183784,
+      "learning_rate": 8.517096748273951e-06,
+      "loss": 0.3624,
+      "step": 2817
+    },
+    {
+      "epoch": 0.27425790754257906,
+      "grad_norm": 1.2633026385457689,
+      "learning_rate": 8.515976281662613e-06,
+      "loss": 0.349,
+      "step": 2818
+    },
+    {
+      "epoch": 0.2743552311435523,
+      "grad_norm": 1.392024932172172,
+      "learning_rate": 8.514855465666546e-06,
+      "loss": 0.4514,
+      "step": 2819
+    },
+    {
+      "epoch": 0.27445255474452557,
+      "grad_norm": 1.4295145565971665,
+      "learning_rate": 8.513734300397135e-06,
+      "loss": 0.5668,
+      "step": 2820
+    },
+    {
+      "epoch": 0.27454987834549877,
+      "grad_norm": 1.0967459926110283,
+      "learning_rate": 8.512612785965787e-06,
+      "loss": 0.1808,
+      "step": 2821
+    },
+    {
+      "epoch": 0.274647201946472,
+      "grad_norm": 1.4843839946273536,
+      "learning_rate": 8.511490922483946e-06,
+      "loss": 0.4352,
+      "step": 2822
+    },
+    {
+      "epoch": 0.2747445255474453,
+      "grad_norm": 1.339649820333997,
+      "learning_rate": 8.510368710063093e-06,
+      "loss": 0.3137,
+      "step": 2823
+    },
+    {
+      "epoch": 0.27484184914841847,
+      "grad_norm": 1.32567882782868,
+      "learning_rate": 8.509246148814745e-06,
+      "loss": 0.4089,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2749391727493917,
+      "grad_norm": 1.2497731956714773,
+      "learning_rate": 8.50812323885045e-06,
+      "loss": 0.382,
+      "step": 2825
+    },
+    {
+      "epoch": 0.275036496350365,
+      "grad_norm": 1.5771259884963846,
+      "learning_rate": 8.506999980281791e-06,
+      "loss": 0.501,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2751338199513382,
+      "grad_norm": 1.3295615561309837,
+      "learning_rate": 8.505876373220393e-06,
+      "loss": 0.3635,
+      "step": 2827
+    },
+    {
+      "epoch": 0.27523114355231143,
+      "grad_norm": 1.55543645713159,
+      "learning_rate": 8.504752417777899e-06,
+      "loss": 0.2986,
+      "step": 2828
+    },
+    {
+      "epoch": 0.2753284671532847,
+      "grad_norm": 1.421283473121396,
+      "learning_rate": 8.503628114066008e-06,
+      "loss": 0.4931,
+      "step": 2829
+    },
+    {
+      "epoch": 0.2754257907542579,
+      "grad_norm": 1.1988827610585986,
+      "learning_rate": 8.502503462196435e-06,
+      "loss": 0.3272,
+      "step": 2830
+    },
+    {
+      "epoch": 0.27552311435523114,
+      "grad_norm": 1.6163491550131937,
+      "learning_rate": 8.501378462280941e-06,
+      "loss": 0.5794,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2756204379562044,
+      "grad_norm": 1.6499795796835799,
+      "learning_rate": 8.500253114431316e-06,
+      "loss": 0.3668,
+      "step": 2832
+    },
+    {
+      "epoch": 0.27571776155717764,
+      "grad_norm": 1.7305434923413188,
+      "learning_rate": 8.499127418759388e-06,
+      "loss": 0.5291,
+      "step": 2833
+    },
+    {
+      "epoch": 0.27581508515815084,
+      "grad_norm": 1.4062980643641485,
+      "learning_rate": 8.498001375377018e-06,
+      "loss": 0.4645,
+      "step": 2834
+    },
+    {
+      "epoch": 0.2759124087591241,
+      "grad_norm": 1.2961260919749351,
+      "learning_rate": 8.496874984396101e-06,
+      "loss": 0.2517,
+      "step": 2835
+    },
+    {
+      "epoch": 0.27600973236009735,
+      "grad_norm": 1.4273972641674804,
+      "learning_rate": 8.495748245928568e-06,
+      "loss": 0.4705,
+      "step": 2836
+    },
+    {
+      "epoch": 0.27610705596107055,
+      "grad_norm": 1.1525746776855315,
+      "learning_rate": 8.494621160086383e-06,
+      "loss": 0.3747,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2762043795620438,
+      "grad_norm": 1.6083708658269757,
+      "learning_rate": 8.493493726981545e-06,
+      "loss": 0.5754,
+      "step": 2838
+    },
+    {
+      "epoch": 0.27630170316301705,
+      "grad_norm": 1.6380932846987073,
+      "learning_rate": 8.492365946726087e-06,
+      "loss": 0.4668,
+      "step": 2839
+    },
+    {
+      "epoch": 0.27639902676399025,
+      "grad_norm": 1.3587028332396105,
+      "learning_rate": 8.491237819432081e-06,
+      "loss": 0.3466,
+      "step": 2840
+    },
+    {
+      "epoch": 0.2764963503649635,
+      "grad_norm": 1.5812508624530597,
+      "learning_rate": 8.490109345211625e-06,
+      "loss": 0.628,
+      "step": 2841
+    },
+    {
+      "epoch": 0.27659367396593676,
+      "grad_norm": 1.359461682943084,
+      "learning_rate": 8.48898052417686e-06,
+      "loss": 0.4799,
+      "step": 2842
+    },
+    {
+      "epoch": 0.27669099756690996,
+      "grad_norm": 1.3773089875645015,
+      "learning_rate": 8.487851356439953e-06,
+      "loss": 0.3064,
+      "step": 2843
+    },
+    {
+      "epoch": 0.2767883211678832,
+      "grad_norm": 1.445505572645753,
+      "learning_rate": 8.486721842113114e-06,
+      "loss": 0.4629,
+      "step": 2844
+    },
+    {
+      "epoch": 0.27688564476885646,
+      "grad_norm": 2.1729540442826796,
+      "learning_rate": 8.485591981308584e-06,
+      "loss": 0.501,
+      "step": 2845
+    },
+    {
+      "epoch": 0.27698296836982966,
+      "grad_norm": 1.2698072866971275,
+      "learning_rate": 8.484461774138635e-06,
+      "loss": 0.3354,
+      "step": 2846
+    },
+    {
+      "epoch": 0.2770802919708029,
+      "grad_norm": 1.2270792461817257,
+      "learning_rate": 8.483331220715578e-06,
+      "loss": 0.2925,
+      "step": 2847
+    },
+    {
+      "epoch": 0.27717761557177617,
+      "grad_norm": 1.4982940191444252,
+      "learning_rate": 8.482200321151757e-06,
+      "loss": 0.4372,
+      "step": 2848
+    },
+    {
+      "epoch": 0.27727493917274937,
+      "grad_norm": 1.7962422459275051,
+      "learning_rate": 8.48106907555955e-06,
+      "loss": 0.2514,
+      "step": 2849
+    },
+    {
+      "epoch": 0.2773722627737226,
+      "grad_norm": 1.1765428275481227,
+      "learning_rate": 8.479937484051368e-06,
+      "loss": 0.2466,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2774695863746959,
+      "grad_norm": 1.3671035304850088,
+      "learning_rate": 8.47880554673966e-06,
+      "loss": 0.4388,
+      "step": 2851
+    },
+    {
+      "epoch": 0.2775669099756691,
+      "grad_norm": 1.584083262413021,
+      "learning_rate": 8.477673263736908e-06,
+      "loss": 0.3117,
+      "step": 2852
+    },
+    {
+      "epoch": 0.2776642335766423,
+      "grad_norm": 1.6251518472003594,
+      "learning_rate": 8.476540635155623e-06,
+      "loss": 0.4661,
+      "step": 2853
+    },
+    {
+      "epoch": 0.2777615571776156,
+      "grad_norm": 1.6392857489539867,
+      "learning_rate": 8.475407661108361e-06,
+      "loss": 0.354,
+      "step": 2854
+    },
+    {
+      "epoch": 0.27785888077858883,
+      "grad_norm": 1.3195625296951223,
+      "learning_rate": 8.474274341707702e-06,
+      "loss": 0.3744,
+      "step": 2855
+    },
+    {
+      "epoch": 0.27795620437956203,
+      "grad_norm": 1.34410915454318,
+      "learning_rate": 8.473140677066267e-06,
+      "loss": 0.4069,
+      "step": 2856
+    },
+    {
+      "epoch": 0.2780535279805353,
+      "grad_norm": 1.0527413957181246,
+      "learning_rate": 8.472006667296709e-06,
+      "loss": 0.2776,
+      "step": 2857
+    },
+    {
+      "epoch": 0.27815085158150854,
+      "grad_norm": 1.496471387248685,
+      "learning_rate": 8.470872312511714e-06,
+      "loss": 0.3642,
+      "step": 2858
+    },
+    {
+      "epoch": 0.27824817518248174,
+      "grad_norm": 1.532429299396127,
+      "learning_rate": 8.469737612824001e-06,
+      "loss": 0.44,
+      "step": 2859
+    },
+    {
+      "epoch": 0.278345498783455,
+      "grad_norm": 1.601112711944827,
+      "learning_rate": 8.468602568346332e-06,
+      "loss": 0.421,
+      "step": 2860
+    },
+    {
+      "epoch": 0.27844282238442825,
+      "grad_norm": 1.5148720198103927,
+      "learning_rate": 8.467467179191493e-06,
+      "loss": 0.5258,
+      "step": 2861
+    },
+    {
+      "epoch": 0.27854014598540144,
+      "grad_norm": 1.573048120862393,
+      "learning_rate": 8.466331445472308e-06,
+      "loss": 0.4507,
+      "step": 2862
+    },
+    {
+      "epoch": 0.2786374695863747,
+      "grad_norm": 1.3938890789758775,
+      "learning_rate": 8.465195367301639e-06,
+      "loss": 0.3365,
+      "step": 2863
+    },
+    {
+      "epoch": 0.27873479318734795,
+      "grad_norm": 1.6895380781567202,
+      "learning_rate": 8.464058944792375e-06,
+      "loss": 0.4132,
+      "step": 2864
+    },
+    {
+      "epoch": 0.27883211678832115,
+      "grad_norm": 1.6880546647255488,
+      "learning_rate": 8.462922178057444e-06,
+      "loss": 0.2605,
+      "step": 2865
+    },
+    {
+      "epoch": 0.2789294403892944,
+      "grad_norm": 1.491755717654464,
+      "learning_rate": 8.46178506720981e-06,
+      "loss": 0.3983,
+      "step": 2866
+    },
+    {
+      "epoch": 0.27902676399026766,
+      "grad_norm": 1.5848666178901887,
+      "learning_rate": 8.460647612362464e-06,
+      "loss": 0.5101,
+      "step": 2867
+    },
+    {
+      "epoch": 0.27912408759124085,
+      "grad_norm": 1.3442317187907376,
+      "learning_rate": 8.459509813628437e-06,
+      "loss": 0.458,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2792214111922141,
+      "grad_norm": 1.8095809186860319,
+      "learning_rate": 8.458371671120795e-06,
+      "loss": 0.382,
+      "step": 2869
+    },
+    {
+      "epoch": 0.27931873479318736,
+      "grad_norm": 0.9909926300929587,
+      "learning_rate": 8.457233184952635e-06,
+      "loss": 0.2292,
+      "step": 2870
+    },
+    {
+      "epoch": 0.27941605839416056,
+      "grad_norm": 1.7013118787018624,
+      "learning_rate": 8.456094355237086e-06,
+      "loss": 0.6861,
+      "step": 2871
+    },
+    {
+      "epoch": 0.2795133819951338,
+      "grad_norm": 3.4293212695090025,
+      "learning_rate": 8.45495518208732e-06,
+      "loss": 0.3233,
+      "step": 2872
+    },
+    {
+      "epoch": 0.27961070559610707,
+      "grad_norm": 1.4903797163776311,
+      "learning_rate": 8.45381566561653e-06,
+      "loss": 0.3231,
+      "step": 2873
+    },
+    {
+      "epoch": 0.27970802919708027,
+      "grad_norm": 1.5615177882070261,
+      "learning_rate": 8.452675805937956e-06,
+      "loss": 0.4125,
+      "step": 2874
+    },
+    {
+      "epoch": 0.2798053527980535,
+      "grad_norm": 1.4099046900170047,
+      "learning_rate": 8.451535603164865e-06,
+      "loss": 0.4967,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2799026763990268,
+      "grad_norm": 1.383217014263479,
+      "learning_rate": 8.450395057410561e-06,
+      "loss": 0.3411,
+      "step": 2876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2661588037606646,
+      "learning_rate": 8.449254168788377e-06,
+      "loss": 0.3734,
+      "step": 2877
+    },
+    {
+      "epoch": 0.2800973236009732,
+      "grad_norm": 1.4107359648240771,
+      "learning_rate": 8.448112937411689e-06,
+      "loss": 0.4765,
+      "step": 2878
+    },
+    {
+      "epoch": 0.2801946472019465,
+      "grad_norm": 1.567373989947911,
+      "learning_rate": 8.446971363393897e-06,
+      "loss": 0.5806,
+      "step": 2879
+    },
+    {
+      "epoch": 0.28029197080291973,
+      "grad_norm": 1.5980994022663064,
+      "learning_rate": 8.445829446848442e-06,
+      "loss": 0.3765,
+      "step": 2880
+    },
+    {
+      "epoch": 0.28038929440389293,
+      "grad_norm": 1.5582627635759285,
+      "learning_rate": 8.444687187888798e-06,
+      "loss": 0.3838,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2804866180048662,
+      "grad_norm": 2.097365147798996,
+      "learning_rate": 8.44354458662847e-06,
+      "loss": 0.6467,
+      "step": 2882
+    },
+    {
+      "epoch": 0.28058394160583944,
+      "grad_norm": 1.5302257615618868,
+      "learning_rate": 8.442401643181e-06,
+      "loss": 0.4415,
+      "step": 2883
+    },
+    {
+      "epoch": 0.28068126520681264,
+      "grad_norm": 1.1646338986978766,
+      "learning_rate": 8.441258357659962e-06,
+      "loss": 0.3176,
+      "step": 2884
+    },
+    {
+      "epoch": 0.2807785888077859,
+      "grad_norm": 1.2287928718701633,
+      "learning_rate": 8.440114730178968e-06,
+      "loss": 0.4175,
+      "step": 2885
+    },
+    {
+      "epoch": 0.28087591240875914,
+      "grad_norm": 1.4416072881006319,
+      "learning_rate": 8.438970760851658e-06,
+      "loss": 0.4838,
+      "step": 2886
+    },
+    {
+      "epoch": 0.28097323600973234,
+      "grad_norm": 1.319870372533973,
+      "learning_rate": 8.437826449791709e-06,
+      "loss": 0.3421,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2810705596107056,
+      "grad_norm": 1.6261475252650914,
+      "learning_rate": 8.436681797112833e-06,
+      "loss": 0.5019,
+      "step": 2888
+    },
+    {
+      "epoch": 0.28116788321167885,
+      "grad_norm": 1.6203143716652342,
+      "learning_rate": 8.435536802928774e-06,
+      "loss": 0.4282,
+      "step": 2889
+    },
+    {
+      "epoch": 0.28126520681265205,
+      "grad_norm": 1.4127079920263665,
+      "learning_rate": 8.434391467353312e-06,
+      "loss": 0.4542,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2813625304136253,
+      "grad_norm": 1.1756885783532405,
+      "learning_rate": 8.433245790500258e-06,
+      "loss": 0.3563,
+      "step": 2891
+    },
+    {
+      "epoch": 0.28145985401459855,
+      "grad_norm": 1.1824997482138238,
+      "learning_rate": 8.43209977248346e-06,
+      "loss": 0.3628,
+      "step": 2892
+    },
+    {
+      "epoch": 0.28155717761557175,
+      "grad_norm": 1.4280724079623635,
+      "learning_rate": 8.430953413416798e-06,
+      "loss": 0.446,
+      "step": 2893
+    },
+    {
+      "epoch": 0.281654501216545,
+      "grad_norm": 1.0710350994410123,
+      "learning_rate": 8.429806713414188e-06,
+      "loss": 0.2016,
+      "step": 2894
+    },
+    {
+      "epoch": 0.28175182481751826,
+      "grad_norm": 1.453985226232095,
+      "learning_rate": 8.428659672589574e-06,
+      "loss": 0.4325,
+      "step": 2895
+    },
+    {
+      "epoch": 0.28184914841849146,
+      "grad_norm": 1.3045306996673216,
+      "learning_rate": 8.427512291056943e-06,
+      "loss": 0.3838,
+      "step": 2896
+    },
+    {
+      "epoch": 0.2819464720194647,
+      "grad_norm": 1.483337521636422,
+      "learning_rate": 8.426364568930309e-06,
+      "loss": 0.4212,
+      "step": 2897
+    },
+    {
+      "epoch": 0.28204379562043796,
+      "grad_norm": 1.0901324802348065,
+      "learning_rate": 8.425216506323721e-06,
+      "loss": 0.2392,
+      "step": 2898
+    },
+    {
+      "epoch": 0.2821411192214112,
+      "grad_norm": 1.3761268679827663,
+      "learning_rate": 8.424068103351264e-06,
+      "loss": 0.4459,
+      "step": 2899
+    },
+    {
+      "epoch": 0.2822384428223844,
+      "grad_norm": 1.461105500215717,
+      "learning_rate": 8.422919360127053e-06,
+      "loss": 0.5018,
+      "step": 2900
+    },
+    {
+      "epoch": 0.28233576642335767,
+      "grad_norm": 1.4314465150478046,
+      "learning_rate": 8.421770276765245e-06,
+      "loss": 0.4474,
+      "step": 2901
+    },
+    {
+      "epoch": 0.2824330900243309,
+      "grad_norm": 1.6060806185106393,
+      "learning_rate": 8.420620853380018e-06,
+      "loss": 0.5798,
+      "step": 2902
+    },
+    {
+      "epoch": 0.2825304136253041,
+      "grad_norm": 1.4468000025910832,
+      "learning_rate": 8.419471090085596e-06,
+      "loss": 0.5597,
+      "step": 2903
+    },
+    {
+      "epoch": 0.2826277372262774,
+      "grad_norm": 5.585104457387235,
+      "learning_rate": 8.41832098699623e-06,
+      "loss": 0.3493,
+      "step": 2904
+    },
+    {
+      "epoch": 0.28272506082725063,
+      "grad_norm": 1.3577816273786794,
+      "learning_rate": 8.417170544226205e-06,
+      "loss": 0.3262,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2828223844282238,
+      "grad_norm": 1.1546363912171016,
+      "learning_rate": 8.416019761889845e-06,
+      "loss": 0.3691,
+      "step": 2906
+    },
+    {
+      "epoch": 0.2829197080291971,
+      "grad_norm": 1.3224407401265832,
+      "learning_rate": 8.4148686401015e-06,
+      "loss": 0.3079,
+      "step": 2907
+    },
+    {
+      "epoch": 0.28301703163017033,
+      "grad_norm": 1.5947860641264806,
+      "learning_rate": 8.413717178975558e-06,
+      "loss": 0.277,
+      "step": 2908
+    },
+    {
+      "epoch": 0.28311435523114353,
+      "grad_norm": 1.343045870800707,
+      "learning_rate": 8.412565378626442e-06,
+      "loss": 0.3448,
+      "step": 2909
+    },
+    {
+      "epoch": 0.2832116788321168,
+      "grad_norm": 1.5567901041780798,
+      "learning_rate": 8.411413239168609e-06,
+      "loss": 0.3954,
+      "step": 2910
+    },
+    {
+      "epoch": 0.28330900243309004,
+      "grad_norm": 1.5232536009297208,
+      "learning_rate": 8.410260760716545e-06,
+      "loss": 0.5103,
+      "step": 2911
+    },
+    {
+      "epoch": 0.28340632603406324,
+      "grad_norm": 1.2493384040941995,
+      "learning_rate": 8.409107943384773e-06,
+      "loss": 0.3671,
+      "step": 2912
+    },
+    {
+      "epoch": 0.2835036496350365,
+      "grad_norm": 1.246217249188392,
+      "learning_rate": 8.407954787287848e-06,
+      "loss": 0.4112,
+      "step": 2913
+    },
+    {
+      "epoch": 0.28360097323600975,
+      "grad_norm": 1.2012340002353967,
+      "learning_rate": 8.406801292540364e-06,
+      "loss": 0.3769,
+      "step": 2914
+    },
+    {
+      "epoch": 0.28369829683698294,
+      "grad_norm": 1.51749407168492,
+      "learning_rate": 8.405647459256939e-06,
+      "loss": 0.5515,
+      "step": 2915
+    },
+    {
+      "epoch": 0.2837956204379562,
+      "grad_norm": 1.1589770762667257,
+      "learning_rate": 8.404493287552232e-06,
+      "loss": 0.2577,
+      "step": 2916
+    },
+    {
+      "epoch": 0.28389294403892945,
+      "grad_norm": 1.5139932402052954,
+      "learning_rate": 8.403338777540936e-06,
+      "loss": 0.4796,
+      "step": 2917
+    },
+    {
+      "epoch": 0.28399026763990265,
+      "grad_norm": 1.5544290759133006,
+      "learning_rate": 8.402183929337774e-06,
+      "loss": 0.4594,
+      "step": 2918
+    },
+    {
+      "epoch": 0.2840875912408759,
+      "grad_norm": 1.3525572627526583,
+      "learning_rate": 8.401028743057503e-06,
+      "loss": 0.3978,
+      "step": 2919
+    },
+    {
+      "epoch": 0.28418491484184916,
+      "grad_norm": 1.3610916698563846,
+      "learning_rate": 8.399873218814916e-06,
+      "loss": 0.4308,
+      "step": 2920
+    },
+    {
+      "epoch": 0.2842822384428224,
+      "grad_norm": 1.2060322500759533,
+      "learning_rate": 8.398717356724837e-06,
+      "loss": 0.482,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2843795620437956,
+      "grad_norm": 1.152727586861314,
+      "learning_rate": 8.397561156902126e-06,
+      "loss": 0.3862,
+      "step": 2922
+    },
+    {
+      "epoch": 0.28447688564476886,
+      "grad_norm": 1.6371195081735355,
+      "learning_rate": 8.396404619461673e-06,
+      "loss": 0.684,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2845742092457421,
+      "grad_norm": 1.4756480619833048,
+      "learning_rate": 8.395247744518407e-06,
+      "loss": 0.4432,
+      "step": 2924
+    },
+    {
+      "epoch": 0.2846715328467153,
+      "grad_norm": 1.3495353534897125,
+      "learning_rate": 8.394090532187286e-06,
+      "loss": 0.4574,
+      "step": 2925
+    },
+    {
+      "epoch": 0.28476885644768857,
+      "grad_norm": 1.361248347874279,
+      "learning_rate": 8.392932982583301e-06,
+      "loss": 0.3117,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2848661800486618,
+      "grad_norm": 1.5493409509214389,
+      "learning_rate": 8.391775095821481e-06,
+      "loss": 0.5949,
+      "step": 2927
+    },
+    {
+      "epoch": 0.284963503649635,
+      "grad_norm": 1.5159424124979992,
+      "learning_rate": 8.390616872016886e-06,
+      "loss": 0.612,
+      "step": 2928
+    },
+    {
+      "epoch": 0.2850608272506083,
+      "grad_norm": 0.9819694068633834,
+      "learning_rate": 8.389458311284606e-06,
+      "loss": 0.2407,
+      "step": 2929
+    },
+    {
+      "epoch": 0.2851581508515815,
+      "grad_norm": 1.4338313849048412,
+      "learning_rate": 8.388299413739772e-06,
+      "loss": 0.504,
+      "step": 2930
+    },
+    {
+      "epoch": 0.2852554744525547,
+      "grad_norm": 1.6033282710660985,
+      "learning_rate": 8.387140179497541e-06,
+      "loss": 0.4686,
+      "step": 2931
+    },
+    {
+      "epoch": 0.285352798053528,
+      "grad_norm": 1.4339139409278308,
+      "learning_rate": 8.38598060867311e-06,
+      "loss": 0.5885,
+      "step": 2932
+    },
+    {
+      "epoch": 0.28545012165450123,
+      "grad_norm": 1.6962944035069916,
+      "learning_rate": 8.384820701381705e-06,
+      "loss": 0.6325,
+      "step": 2933
+    },
+    {
+      "epoch": 0.28554744525547443,
+      "grad_norm": 1.2380931242026982,
+      "learning_rate": 8.383660457738585e-06,
+      "loss": 0.3528,
+      "step": 2934
+    },
+    {
+      "epoch": 0.2856447688564477,
+      "grad_norm": 1.4958548492045998,
+      "learning_rate": 8.382499877859046e-06,
+      "loss": 0.5261,
+      "step": 2935
+    },
+    {
+      "epoch": 0.28574209245742094,
+      "grad_norm": 1.2493863506860636,
+      "learning_rate": 8.381338961858417e-06,
+      "loss": 0.295,
+      "step": 2936
+    },
+    {
+      "epoch": 0.28583941605839414,
+      "grad_norm": 1.0264542939220365,
+      "learning_rate": 8.380177709852055e-06,
+      "loss": 0.2736,
+      "step": 2937
+    },
+    {
+      "epoch": 0.2859367396593674,
+      "grad_norm": 1.3694559515073481,
+      "learning_rate": 8.379016121955358e-06,
+      "loss": 0.2437,
+      "step": 2938
+    },
+    {
+      "epoch": 0.28603406326034064,
+      "grad_norm": 1.3958652644514353,
+      "learning_rate": 8.377854198283751e-06,
+      "loss": 0.5162,
+      "step": 2939
+    },
+    {
+      "epoch": 0.28613138686131384,
+      "grad_norm": 1.3188642877167738,
+      "learning_rate": 8.376691938952694e-06,
+      "loss": 0.4403,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2862287104622871,
+      "grad_norm": 1.5563883463328907,
+      "learning_rate": 8.375529344077686e-06,
+      "loss": 0.3871,
+      "step": 2941
+    },
+    {
+      "epoch": 0.28632603406326035,
+      "grad_norm": 1.7106139691477682,
+      "learning_rate": 8.37436641377425e-06,
+      "loss": 0.5998,
+      "step": 2942
+    },
+    {
+      "epoch": 0.2864233576642336,
+      "grad_norm": 1.8227768617334648,
+      "learning_rate": 8.373203148157953e-06,
+      "loss": 0.4192,
+      "step": 2943
+    },
+    {
+      "epoch": 0.2865206812652068,
+      "grad_norm": 1.3645142496496503,
+      "learning_rate": 8.372039547344383e-06,
+      "loss": 0.4301,
+      "step": 2944
+    },
+    {
+      "epoch": 0.28661800486618005,
+      "grad_norm": 1.4644520960794265,
+      "learning_rate": 8.370875611449173e-06,
+      "loss": 0.4333,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2867153284671533,
+      "grad_norm": 1.3686778637415178,
+      "learning_rate": 8.369711340587981e-06,
+      "loss": 0.4735,
+      "step": 2946
+    },
+    {
+      "epoch": 0.2868126520681265,
+      "grad_norm": 1.7752150982830557,
+      "learning_rate": 8.368546734876499e-06,
+      "loss": 0.605,
+      "step": 2947
+    },
+    {
+      "epoch": 0.28690997566909976,
+      "grad_norm": 1.6349896239905135,
+      "learning_rate": 8.36738179443046e-06,
+      "loss": 0.4521,
+      "step": 2948
+    },
+    {
+      "epoch": 0.287007299270073,
+      "grad_norm": 1.7001103309282906,
+      "learning_rate": 8.366216519365623e-06,
+      "loss": 0.5243,
+      "step": 2949
+    },
+    {
+      "epoch": 0.2871046228710462,
+      "grad_norm": 1.3288526449094853,
+      "learning_rate": 8.365050909797779e-06,
+      "loss": 0.4226,
+      "step": 2950
+    },
+    {
+      "epoch": 0.28720194647201946,
+      "grad_norm": 1.0609308885865543,
+      "learning_rate": 8.36388496584276e-06,
+      "loss": 0.2761,
+      "step": 2951
+    },
+    {
+      "epoch": 0.2872992700729927,
+      "grad_norm": 1.3048762567541314,
+      "learning_rate": 8.362718687616422e-06,
+      "loss": 0.3166,
+      "step": 2952
+    },
+    {
+      "epoch": 0.2873965936739659,
+      "grad_norm": 1.5602591658770568,
+      "learning_rate": 8.361552075234664e-06,
+      "loss": 0.1814,
+      "step": 2953
+    },
+    {
+      "epoch": 0.28749391727493917,
+      "grad_norm": 1.261612878851385,
+      "learning_rate": 8.360385128813409e-06,
+      "loss": 0.3431,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2875912408759124,
+      "grad_norm": 1.6502840086679433,
+      "learning_rate": 8.359217848468617e-06,
+      "loss": 0.5688,
+      "step": 2955
+    },
+    {
+      "epoch": 0.2876885644768856,
+      "grad_norm": 1.1758618501430975,
+      "learning_rate": 8.358050234316283e-06,
+      "loss": 0.376,
+      "step": 2956
+    },
+    {
+      "epoch": 0.2877858880778589,
+      "grad_norm": 1.3748216513361973,
+      "learning_rate": 8.356882286472433e-06,
+      "loss": 0.4893,
+      "step": 2957
+    },
+    {
+      "epoch": 0.28788321167883213,
+      "grad_norm": 1.490557754247365,
+      "learning_rate": 8.35571400505313e-06,
+      "loss": 0.4322,
+      "step": 2958
+    },
+    {
+      "epoch": 0.2879805352798053,
+      "grad_norm": 1.2474734521766377,
+      "learning_rate": 8.35454539017446e-06,
+      "loss": 0.249,
+      "step": 2959
+    },
+    {
+      "epoch": 0.2880778588807786,
+      "grad_norm": 1.3041956082790018,
+      "learning_rate": 8.353376441952554e-06,
+      "loss": 0.3629,
+      "step": 2960
+    },
+    {
+      "epoch": 0.28817518248175183,
+      "grad_norm": 1.1813542799359134,
+      "learning_rate": 8.352207160503572e-06,
+      "loss": 0.2541,
+      "step": 2961
+    },
+    {
+      "epoch": 0.2882725060827251,
+      "grad_norm": 1.6196703441196314,
+      "learning_rate": 8.351037545943702e-06,
+      "loss": 0.5863,
+      "step": 2962
+    },
+    {
+      "epoch": 0.2883698296836983,
+      "grad_norm": 1.6020435634219072,
+      "learning_rate": 8.34986759838917e-06,
+      "loss": 0.5539,
+      "step": 2963
+    },
+    {
+      "epoch": 0.28846715328467154,
+      "grad_norm": 1.6170521555116952,
+      "learning_rate": 8.348697317956238e-06,
+      "loss": 0.4234,
+      "step": 2964
+    },
+    {
+      "epoch": 0.2885644768856448,
+      "grad_norm": 1.2300623631368495,
+      "learning_rate": 8.347526704761193e-06,
+      "loss": 0.2784,
+      "step": 2965
+    },
+    {
+      "epoch": 0.288661800486618,
+      "grad_norm": 2.179168092375873,
+      "learning_rate": 8.346355758920364e-06,
+      "loss": 0.4561,
+      "step": 2966
+    },
+    {
+      "epoch": 0.28875912408759125,
+      "grad_norm": 1.5135423174141494,
+      "learning_rate": 8.345184480550104e-06,
+      "loss": 0.3807,
+      "step": 2967
+    },
+    {
+      "epoch": 0.2888564476885645,
+      "grad_norm": 1.7005351963186346,
+      "learning_rate": 8.344012869766808e-06,
+      "loss": 0.538,
+      "step": 2968
+    },
+    {
+      "epoch": 0.2889537712895377,
+      "grad_norm": 1.2789157911351394,
+      "learning_rate": 8.342840926686898e-06,
+      "loss": 0.2623,
+      "step": 2969
+    },
+    {
+      "epoch": 0.28905109489051095,
+      "grad_norm": 1.304761873055631,
+      "learning_rate": 8.34166865142683e-06,
+      "loss": 0.4219,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2891484184914842,
+      "grad_norm": 1.6192760894025877,
+      "learning_rate": 8.340496044103095e-06,
+      "loss": 0.4378,
+      "step": 2971
+    },
+    {
+      "epoch": 0.2892457420924574,
+      "grad_norm": 1.4363442626245757,
+      "learning_rate": 8.339323104832214e-06,
+      "loss": 0.3819,
+      "step": 2972
+    },
+    {
+      "epoch": 0.28934306569343066,
+      "grad_norm": 1.5094300127764981,
+      "learning_rate": 8.338149833730742e-06,
+      "loss": 0.2769,
+      "step": 2973
+    },
+    {
+      "epoch": 0.2894403892944039,
+      "grad_norm": 1.6047897202306092,
+      "learning_rate": 8.33697623091527e-06,
+      "loss": 0.424,
+      "step": 2974
+    },
+    {
+      "epoch": 0.2895377128953771,
+      "grad_norm": 1.3129110600868221,
+      "learning_rate": 8.33580229650242e-06,
+      "loss": 0.5053,
+      "step": 2975
+    },
+    {
+      "epoch": 0.28963503649635036,
+      "grad_norm": 1.1812562932245452,
+      "learning_rate": 8.334628030608845e-06,
+      "loss": 0.3835,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2897323600973236,
+      "grad_norm": 1.2211203388582414,
+      "learning_rate": 8.333453433351233e-06,
+      "loss": 0.3531,
+      "step": 2977
+    },
+    {
+      "epoch": 0.2898296836982968,
+      "grad_norm": 1.4620903484748373,
+      "learning_rate": 8.332278504846303e-06,
+      "loss": 0.4771,
+      "step": 2978
+    },
+    {
+      "epoch": 0.28992700729927007,
+      "grad_norm": 0.9704255718501243,
+      "learning_rate": 8.331103245210812e-06,
+      "loss": 0.2618,
+      "step": 2979
+    },
+    {
+      "epoch": 0.2900243309002433,
+      "grad_norm": 1.2827724622455963,
+      "learning_rate": 8.329927654561544e-06,
+      "loss": 0.3052,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2901216545012165,
+      "grad_norm": 1.378581411338256,
+      "learning_rate": 8.328751733015316e-06,
+      "loss": 0.3568,
+      "step": 2981
+    },
+    {
+      "epoch": 0.2902189781021898,
+      "grad_norm": 1.769807570821765,
+      "learning_rate": 8.327575480688985e-06,
+      "loss": 0.3102,
+      "step": 2982
+    },
+    {
+      "epoch": 0.290316301703163,
+      "grad_norm": 1.4326301683333176,
+      "learning_rate": 8.32639889769943e-06,
+      "loss": 0.3218,
+      "step": 2983
+    },
+    {
+      "epoch": 0.2904136253041363,
+      "grad_norm": 1.5418816322088151,
+      "learning_rate": 8.325221984163575e-06,
+      "loss": 0.3257,
+      "step": 2984
+    },
+    {
+      "epoch": 0.2905109489051095,
+      "grad_norm": 1.573484642436306,
+      "learning_rate": 8.324044740198366e-06,
+      "loss": 0.5401,
+      "step": 2985
+    },
+    {
+      "epoch": 0.29060827250608273,
+      "grad_norm": 1.2270555416429247,
+      "learning_rate": 8.322867165920789e-06,
+      "loss": 0.3914,
+      "step": 2986
+    },
+    {
+      "epoch": 0.290705596107056,
+      "grad_norm": 1.1838846887742434,
+      "learning_rate": 8.321689261447858e-06,
+      "loss": 0.3282,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2908029197080292,
+      "grad_norm": 1.5077214188811954,
+      "learning_rate": 8.320511026896624e-06,
+      "loss": 0.5279,
+      "step": 2988
+    },
+    {
+      "epoch": 0.29090024330900244,
+      "grad_norm": 1.1784061774291985,
+      "learning_rate": 8.31933246238417e-06,
+      "loss": 0.403,
+      "step": 2989
+    },
+    {
+      "epoch": 0.2909975669099757,
+      "grad_norm": 1.2176703537151474,
+      "learning_rate": 8.318153568027607e-06,
+      "loss": 0.4213,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2910948905109489,
+      "grad_norm": 1.3475262123063816,
+      "learning_rate": 8.316974343944085e-06,
+      "loss": 0.4059,
+      "step": 2991
+    },
+    {
+      "epoch": 0.29119221411192214,
+      "grad_norm": 1.2398233047847593,
+      "learning_rate": 8.315794790250784e-06,
+      "loss": 0.2626,
+      "step": 2992
+    },
+    {
+      "epoch": 0.2912895377128954,
+      "grad_norm": 1.3862498175549538,
+      "learning_rate": 8.314614907064915e-06,
+      "loss": 0.4535,
+      "step": 2993
+    },
+    {
+      "epoch": 0.2913868613138686,
+      "grad_norm": 1.455622096437578,
+      "learning_rate": 8.313434694503727e-06,
+      "loss": 0.4067,
+      "step": 2994
+    },
+    {
+      "epoch": 0.29148418491484185,
+      "grad_norm": 1.4755183973829757,
+      "learning_rate": 8.312254152684496e-06,
+      "loss": 0.6493,
+      "step": 2995
+    },
+    {
+      "epoch": 0.2915815085158151,
+      "grad_norm": 1.0399713771806027,
+      "learning_rate": 8.311073281724536e-06,
+      "loss": 0.3051,
+      "step": 2996
+    },
+    {
+      "epoch": 0.2916788321167883,
+      "grad_norm": 1.3151300509583979,
+      "learning_rate": 8.309892081741186e-06,
+      "loss": 0.3982,
+      "step": 2997
+    },
+    {
+      "epoch": 0.29177615571776155,
+      "grad_norm": 1.376541833798208,
+      "learning_rate": 8.308710552851826e-06,
+      "loss": 0.4749,
+      "step": 2998
+    },
+    {
+      "epoch": 0.2918734793187348,
+      "grad_norm": 1.2551786912554768,
+      "learning_rate": 8.307528695173865e-06,
+      "loss": 0.3118,
+      "step": 2999
+    },
+    {
+      "epoch": 0.291970802919708,
+      "grad_norm": 2.1707038191553463,
+      "learning_rate": 8.306346508824746e-06,
+      "loss": 0.3438,
+      "step": 3000
+    },
+    {
+      "epoch": 0.29206812652068126,
+      "grad_norm": 1.4299459588569998,
+      "learning_rate": 8.30516399392194e-06,
+      "loss": 0.4838,
+      "step": 3001
+    },
+    {
+      "epoch": 0.2921654501216545,
+      "grad_norm": 1.378341342959643,
+      "learning_rate": 8.303981150582958e-06,
+      "loss": 0.5055,
+      "step": 3002
+    },
+    {
+      "epoch": 0.2922627737226277,
+      "grad_norm": 1.4826508798742193,
+      "learning_rate": 8.302797978925338e-06,
+      "loss": 0.3737,
+      "step": 3003
+    },
+    {
+      "epoch": 0.29236009732360096,
+      "grad_norm": 1.222513403789782,
+      "learning_rate": 8.301614479066653e-06,
+      "loss": 0.4587,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2924574209245742,
+      "grad_norm": 1.3819233250029228,
+      "learning_rate": 8.300430651124508e-06,
+      "loss": 0.4021,
+      "step": 3005
+    },
+    {
+      "epoch": 0.29255474452554747,
+      "grad_norm": 1.2846536784172882,
+      "learning_rate": 8.29924649521654e-06,
+      "loss": 0.3609,
+      "step": 3006
+    },
+    {
+      "epoch": 0.29265206812652067,
+      "grad_norm": 1.4274226525457885,
+      "learning_rate": 8.298062011460419e-06,
+      "loss": 0.5267,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2927493917274939,
+      "grad_norm": 1.4642655922839627,
+      "learning_rate": 8.296877199973849e-06,
+      "loss": 0.3499,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2928467153284672,
+      "grad_norm": 1.4317302181421974,
+      "learning_rate": 8.295692060874568e-06,
+      "loss": 0.4979,
+      "step": 3009
+    },
+    {
+      "epoch": 0.2929440389294404,
+      "grad_norm": 1.3191877461185262,
+      "learning_rate": 8.294506594280338e-06,
+      "loss": 0.2835,
+      "step": 3010
+    },
+    {
+      "epoch": 0.29304136253041363,
+      "grad_norm": 1.0943861065294986,
+      "learning_rate": 8.293320800308964e-06,
+      "loss": 0.2138,
+      "step": 3011
+    },
+    {
+      "epoch": 0.2931386861313869,
+      "grad_norm": 1.2621219805575281,
+      "learning_rate": 8.292134679078277e-06,
+      "loss": 0.3027,
+      "step": 3012
+    },
+    {
+      "epoch": 0.2932360097323601,
+      "grad_norm": 1.556172337566686,
+      "learning_rate": 8.290948230706145e-06,
+      "loss": 0.4462,
+      "step": 3013
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 1.3363658374504028,
+      "learning_rate": 8.289761455310463e-06,
+      "loss": 0.373,
+      "step": 3014
+    },
+    {
+      "epoch": 0.2934306569343066,
+      "grad_norm": 1.4458593210455408,
+      "learning_rate": 8.288574353009164e-06,
+      "loss": 0.5566,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2935279805352798,
+      "grad_norm": 1.5034274044899172,
+      "learning_rate": 8.287386923920211e-06,
+      "loss": 0.3837,
+      "step": 3016
+    },
+    {
+      "epoch": 0.29362530413625304,
+      "grad_norm": 1.484769748600726,
+      "learning_rate": 8.286199168161598e-06,
+      "loss": 0.3173,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2937226277372263,
+      "grad_norm": 1.4336064725306121,
+      "learning_rate": 8.285011085851353e-06,
+      "loss": 0.4005,
+      "step": 3018
+    },
+    {
+      "epoch": 0.2938199513381995,
+      "grad_norm": 1.3857231757141482,
+      "learning_rate": 8.283822677107539e-06,
+      "loss": 0.481,
+      "step": 3019
+    },
+    {
+      "epoch": 0.29391727493917275,
+      "grad_norm": 1.4086307294395457,
+      "learning_rate": 8.282633942048244e-06,
+      "loss": 0.4181,
+      "step": 3020
+    },
+    {
+      "epoch": 0.294014598540146,
+      "grad_norm": 1.4701075671537391,
+      "learning_rate": 8.2814448807916e-06,
+      "loss": 0.4041,
+      "step": 3021
+    },
+    {
+      "epoch": 0.2941119221411192,
+      "grad_norm": 1.5925621393078395,
+      "learning_rate": 8.28025549345576e-06,
+      "loss": 0.3062,
+      "step": 3022
+    },
+    {
+      "epoch": 0.29420924574209245,
+      "grad_norm": 1.6058911141553376,
+      "learning_rate": 8.279065780158914e-06,
+      "loss": 0.5534,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2943065693430657,
+      "grad_norm": 1.4134575830281486,
+      "learning_rate": 8.277875741019289e-06,
+      "loss": 0.5017,
+      "step": 3024
+    },
+    {
+      "epoch": 0.2944038929440389,
+      "grad_norm": 1.6163740830610969,
+      "learning_rate": 8.276685376155133e-06,
+      "loss": 0.5513,
+      "step": 3025
+    },
+    {
+      "epoch": 0.29450121654501216,
+      "grad_norm": 1.3415920762045879,
+      "learning_rate": 8.275494685684739e-06,
+      "loss": 0.4209,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2945985401459854,
+      "grad_norm": 1.699522776097275,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 0.2444,
+      "step": 3027
+    },
+    {
+      "epoch": 0.29469586374695866,
+      "grad_norm": 1.3118143561432465,
+      "learning_rate": 8.273112328398545e-06,
+      "loss": 0.3282,
+      "step": 3028
+    },
+    {
+      "epoch": 0.29479318734793186,
+      "grad_norm": 1.3608335365502384,
+      "learning_rate": 8.271920661819479e-06,
+      "loss": 0.4625,
+      "step": 3029
+    },
+    {
+      "epoch": 0.2948905109489051,
+      "grad_norm": 1.320965035708582,
+      "learning_rate": 8.270728670107645e-06,
+      "loss": 0.4161,
+      "step": 3030
+    },
+    {
+      "epoch": 0.29498783454987837,
+      "grad_norm": 1.2315684415049128,
+      "learning_rate": 8.269536353381493e-06,
+      "loss": 0.3264,
+      "step": 3031
+    },
+    {
+      "epoch": 0.29508515815085157,
+      "grad_norm": 1.2397754210481065,
+      "learning_rate": 8.268343711759505e-06,
+      "loss": 0.3184,
+      "step": 3032
+    },
+    {
+      "epoch": 0.2951824817518248,
+      "grad_norm": 1.4717261820272485,
+      "learning_rate": 8.267150745360194e-06,
+      "loss": 0.381,
+      "step": 3033
+    },
+    {
+      "epoch": 0.2952798053527981,
+      "grad_norm": 1.7364842416546407,
+      "learning_rate": 8.265957454302102e-06,
+      "loss": 0.3639,
+      "step": 3034
+    },
+    {
+      "epoch": 0.2953771289537713,
+      "grad_norm": 1.6249980192905973,
+      "learning_rate": 8.264763838703813e-06,
+      "loss": 0.5112,
+      "step": 3035
+    },
+    {
+      "epoch": 0.2954744525547445,
+      "grad_norm": 1.682249094263979,
+      "learning_rate": 8.263569898683934e-06,
+      "loss": 0.4894,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2955717761557178,
+      "grad_norm": 1.9200248186176307,
+      "learning_rate": 8.262375634361108e-06,
+      "loss": 0.529,
+      "step": 3037
+    },
+    {
+      "epoch": 0.295669099756691,
+      "grad_norm": 1.4426650259998133,
+      "learning_rate": 8.261181045854011e-06,
+      "loss": 0.5037,
+      "step": 3038
+    },
+    {
+      "epoch": 0.29576642335766423,
+      "grad_norm": 1.6904227765149746,
+      "learning_rate": 8.259986133281348e-06,
+      "loss": 0.3632,
+      "step": 3039
+    },
+    {
+      "epoch": 0.2958637469586375,
+      "grad_norm": 1.3863799205056755,
+      "learning_rate": 8.25879089676186e-06,
+      "loss": 0.4148,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2959610705596107,
+      "grad_norm": 1.627436205526306,
+      "learning_rate": 8.257595336414317e-06,
+      "loss": 0.4558,
+      "step": 3041
+    },
+    {
+      "epoch": 0.29605839416058394,
+      "grad_norm": 1.3163567598814478,
+      "learning_rate": 8.256399452357524e-06,
+      "loss": 0.2713,
+      "step": 3042
+    },
+    {
+      "epoch": 0.2961557177615572,
+      "grad_norm": 1.6072179171276018,
+      "learning_rate": 8.255203244710316e-06,
+      "loss": 0.353,
+      "step": 3043
+    },
+    {
+      "epoch": 0.2962530413625304,
+      "grad_norm": 1.4217719575203627,
+      "learning_rate": 8.254006713591559e-06,
+      "loss": 0.3744,
+      "step": 3044
+    },
+    {
+      "epoch": 0.29635036496350364,
+      "grad_norm": 1.9013012922141048,
+      "learning_rate": 8.252809859120154e-06,
+      "loss": 0.209,
+      "step": 3045
+    },
+    {
+      "epoch": 0.2964476885644769,
+      "grad_norm": 1.390657831725977,
+      "learning_rate": 8.251612681415035e-06,
+      "loss": 0.3722,
+      "step": 3046
+    },
+    {
+      "epoch": 0.2965450121654501,
+      "grad_norm": 1.4478686848472833,
+      "learning_rate": 8.250415180595167e-06,
+      "loss": 0.3869,
+      "step": 3047
+    },
+    {
+      "epoch": 0.29664233576642335,
+      "grad_norm": 1.1443911522017596,
+      "learning_rate": 8.249217356779544e-06,
+      "loss": 0.3385,
+      "step": 3048
+    },
+    {
+      "epoch": 0.2967396593673966,
+      "grad_norm": 1.7245119786652503,
+      "learning_rate": 8.248019210087195e-06,
+      "loss": 0.3023,
+      "step": 3049
+    },
+    {
+      "epoch": 0.29683698296836986,
+      "grad_norm": 1.8030337728763741,
+      "learning_rate": 8.24682074063718e-06,
+      "loss": 0.3784,
+      "step": 3050
+    },
+    {
+      "epoch": 0.29693430656934305,
+      "grad_norm": 1.299417141317702,
+      "learning_rate": 8.245621948548593e-06,
+      "loss": 0.2963,
+      "step": 3051
+    },
+    {
+      "epoch": 0.2970316301703163,
+      "grad_norm": 1.3334468356141627,
+      "learning_rate": 8.244422833940558e-06,
+      "loss": 0.3671,
+      "step": 3052
+    },
+    {
+      "epoch": 0.29712895377128956,
+      "grad_norm": 1.6168488226188178,
+      "learning_rate": 8.24322339693223e-06,
+      "loss": 0.5497,
+      "step": 3053
+    },
+    {
+      "epoch": 0.29722627737226276,
+      "grad_norm": 1.49700230831562,
+      "learning_rate": 8.242023637642802e-06,
+      "loss": 0.4567,
+      "step": 3054
+    },
+    {
+      "epoch": 0.297323600973236,
+      "grad_norm": 1.0494586888942983,
+      "learning_rate": 8.24082355619149e-06,
+      "loss": 0.2186,
+      "step": 3055
+    },
+    {
+      "epoch": 0.29742092457420927,
+      "grad_norm": 1.372792205417397,
+      "learning_rate": 8.239623152697553e-06,
+      "loss": 0.5083,
+      "step": 3056
+    },
+    {
+      "epoch": 0.29751824817518246,
+      "grad_norm": 1.266230497219453,
+      "learning_rate": 8.238422427280269e-06,
+      "loss": 0.461,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2976155717761557,
+      "grad_norm": 1.5041389582539588,
+      "learning_rate": 8.237221380058959e-06,
+      "loss": 0.3813,
+      "step": 3058
+    },
+    {
+      "epoch": 0.29771289537712897,
+      "grad_norm": 1.4593593621079823,
+      "learning_rate": 8.23602001115297e-06,
+      "loss": 0.473,
+      "step": 3059
+    },
+    {
+      "epoch": 0.29781021897810217,
+      "grad_norm": 1.3666083716931219,
+      "learning_rate": 8.234818320681685e-06,
+      "loss": 0.4822,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2979075425790754,
+      "grad_norm": 1.407870228183954,
+      "learning_rate": 8.233616308764513e-06,
+      "loss": 0.4012,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2980048661800487,
+      "grad_norm": 1.4404350668596586,
+      "learning_rate": 8.232413975520903e-06,
+      "loss": 0.5057,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2981021897810219,
+      "grad_norm": 1.3912456713229528,
+      "learning_rate": 8.231211321070329e-06,
+      "loss": 0.4578,
+      "step": 3063
+    },
+    {
+      "epoch": 0.29819951338199513,
+      "grad_norm": 1.3191795228165797,
+      "learning_rate": 8.2300083455323e-06,
+      "loss": 0.3888,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2982968369829684,
+      "grad_norm": 1.4258248936492355,
+      "learning_rate": 8.228805049026355e-06,
+      "loss": 0.5108,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2983941605839416,
+      "grad_norm": 1.4850835614825084,
+      "learning_rate": 8.22760143167207e-06,
+      "loss": 0.5968,
+      "step": 3066
+    },
+    {
+      "epoch": 0.29849148418491483,
+      "grad_norm": 1.2696050534436827,
+      "learning_rate": 8.226397493589044e-06,
+      "loss": 0.3328,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2985888077858881,
+      "grad_norm": 1.1993181516723008,
+      "learning_rate": 8.225193234896918e-06,
+      "loss": 0.2682,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2986861313868613,
+      "grad_norm": 1.3420953543565923,
+      "learning_rate": 8.223988655715355e-06,
+      "loss": 0.3865,
+      "step": 3069
+    },
+    {
+      "epoch": 0.29878345498783454,
+      "grad_norm": 1.305913976862295,
+      "learning_rate": 8.222783756164061e-06,
+      "loss": 0.3551,
+      "step": 3070
+    },
+    {
+      "epoch": 0.2988807785888078,
+      "grad_norm": 1.3385899852932626,
+      "learning_rate": 8.221578536362764e-06,
+      "loss": 0.4203,
+      "step": 3071
+    },
+    {
+      "epoch": 0.29897810218978105,
+      "grad_norm": 1.189534251886867,
+      "learning_rate": 8.220372996431228e-06,
+      "loss": 0.2937,
+      "step": 3072
+    },
+    {
+      "epoch": 0.29907542579075425,
+      "grad_norm": 1.5982329206910104,
+      "learning_rate": 8.219167136489245e-06,
+      "loss": 0.6064,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2991727493917275,
+      "grad_norm": 1.775024980718492,
+      "learning_rate": 8.217960956656648e-06,
+      "loss": 0.5517,
+      "step": 3074
+    },
+    {
+      "epoch": 0.29927007299270075,
+      "grad_norm": 1.4818012612095348,
+      "learning_rate": 8.216754457053291e-06,
+      "loss": 0.3574,
+      "step": 3075
+    },
+    {
+      "epoch": 0.29936739659367395,
+      "grad_norm": 1.5621403089409462,
+      "learning_rate": 8.215547637799068e-06,
+      "loss": 0.4108,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2994647201946472,
+      "grad_norm": 1.4983847186167278,
+      "learning_rate": 8.214340499013899e-06,
+      "loss": 0.4644,
+      "step": 3077
+    },
+    {
+      "epoch": 0.29956204379562046,
+      "grad_norm": 1.5897848132407382,
+      "learning_rate": 8.213133040817738e-06,
+      "loss": 0.4894,
+      "step": 3078
+    },
+    {
+      "epoch": 0.29965936739659366,
+      "grad_norm": 1.6354640621760643,
+      "learning_rate": 8.211925263330573e-06,
+      "loss": 0.4583,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2997566909975669,
+      "grad_norm": 1.4952024987397354,
+      "learning_rate": 8.21071716667242e-06,
+      "loss": 0.5976,
+      "step": 3080
+    },
+    {
+      "epoch": 0.29985401459854016,
+      "grad_norm": 1.0095340308225043,
+      "learning_rate": 8.20950875096333e-06,
+      "loss": 0.2524,
+      "step": 3081
+    },
+    {
+      "epoch": 0.29995133819951336,
+      "grad_norm": 1.4197678935056404,
+      "learning_rate": 8.208300016323381e-06,
+      "loss": 0.5514,
+      "step": 3082
+    },
+    {
+      "epoch": 0.3000486618004866,
+      "grad_norm": 1.249287306745543,
+      "learning_rate": 8.207090962872688e-06,
+      "loss": 0.2683,
+      "step": 3083
+    },
+    {
+      "epoch": 0.30014598540145987,
+      "grad_norm": 1.2420194980085992,
+      "learning_rate": 8.205881590731394e-06,
+      "loss": 0.3941,
+      "step": 3084
+    },
+    {
+      "epoch": 0.30024330900243307,
+      "grad_norm": 1.0228818593574307,
+      "learning_rate": 8.204671900019676e-06,
+      "loss": 0.2158,
+      "step": 3085
+    },
+    {
+      "epoch": 0.3003406326034063,
+      "grad_norm": 1.4988207950368069,
+      "learning_rate": 8.203461890857743e-06,
+      "loss": 0.4833,
+      "step": 3086
+    },
+    {
+      "epoch": 0.3004379562043796,
+      "grad_norm": 1.3402746636459373,
+      "learning_rate": 8.20225156336583e-06,
+      "loss": 0.437,
+      "step": 3087
+    },
+    {
+      "epoch": 0.3005352798053528,
+      "grad_norm": 1.3071666622302105,
+      "learning_rate": 8.201040917664214e-06,
+      "loss": 0.3667,
+      "step": 3088
+    },
+    {
+      "epoch": 0.300632603406326,
+      "grad_norm": 2.001934665501785,
+      "learning_rate": 8.199829953873192e-06,
+      "loss": 0.346,
+      "step": 3089
+    },
+    {
+      "epoch": 0.3007299270072993,
+      "grad_norm": 1.50451394225963,
+      "learning_rate": 8.198618672113104e-06,
+      "loss": 0.4897,
+      "step": 3090
+    },
+    {
+      "epoch": 0.3008272506082725,
+      "grad_norm": 1.5127622960173581,
+      "learning_rate": 8.197407072504309e-06,
+      "loss": 0.4301,
+      "step": 3091
+    },
+    {
+      "epoch": 0.30092457420924573,
+      "grad_norm": 1.409495275402236,
+      "learning_rate": 8.196195155167211e-06,
+      "loss": 0.4954,
+      "step": 3092
+    },
+    {
+      "epoch": 0.301021897810219,
+      "grad_norm": 1.3458224438835962,
+      "learning_rate": 8.194982920222233e-06,
+      "loss": 0.5023,
+      "step": 3093
+    },
+    {
+      "epoch": 0.30111922141119224,
+      "grad_norm": 1.484836707336815,
+      "learning_rate": 8.19377036778984e-06,
+      "loss": 0.4471,
+      "step": 3094
+    },
+    {
+      "epoch": 0.30121654501216544,
+      "grad_norm": 1.4314600061658445,
+      "learning_rate": 8.192557497990522e-06,
+      "loss": 0.4519,
+      "step": 3095
+    },
+    {
+      "epoch": 0.3013138686131387,
+      "grad_norm": 1.228152077257465,
+      "learning_rate": 8.191344310944803e-06,
+      "loss": 0.2338,
+      "step": 3096
+    },
+    {
+      "epoch": 0.30141119221411194,
+      "grad_norm": 1.4025619039626473,
+      "learning_rate": 8.19013080677324e-06,
+      "loss": 0.3748,
+      "step": 3097
+    },
+    {
+      "epoch": 0.30150851581508514,
+      "grad_norm": 1.535338102251852,
+      "learning_rate": 8.188916985596415e-06,
+      "loss": 0.3129,
+      "step": 3098
+    },
+    {
+      "epoch": 0.3016058394160584,
+      "grad_norm": 1.7024230210298346,
+      "learning_rate": 8.187702847534952e-06,
+      "loss": 0.5525,
+      "step": 3099
+    },
+    {
+      "epoch": 0.30170316301703165,
+      "grad_norm": 1.4950690283515784,
+      "learning_rate": 8.186488392709495e-06,
+      "loss": 0.5258,
+      "step": 3100
+    },
+    {
+      "epoch": 0.30180048661800485,
+      "grad_norm": 1.589216178732189,
+      "learning_rate": 8.18527362124073e-06,
+      "loss": 0.5745,
+      "step": 3101
+    },
+    {
+      "epoch": 0.3018978102189781,
+      "grad_norm": 1.5942675928105552,
+      "learning_rate": 8.184058533249367e-06,
+      "loss": 0.6344,
+      "step": 3102
+    },
+    {
+      "epoch": 0.30199513381995136,
+      "grad_norm": 1.3981131065521017,
+      "learning_rate": 8.18284312885615e-06,
+      "loss": 0.3369,
+      "step": 3103
+    },
+    {
+      "epoch": 0.30209245742092455,
+      "grad_norm": 1.6180199585993311,
+      "learning_rate": 8.181627408181854e-06,
+      "loss": 0.4014,
+      "step": 3104
+    },
+    {
+      "epoch": 0.3021897810218978,
+      "grad_norm": 1.6338683004824879,
+      "learning_rate": 8.180411371347288e-06,
+      "loss": 0.4983,
+      "step": 3105
+    },
+    {
+      "epoch": 0.30228710462287106,
+      "grad_norm": 1.5225224020676915,
+      "learning_rate": 8.17919501847329e-06,
+      "loss": 0.5016,
+      "step": 3106
+    },
+    {
+      "epoch": 0.30238442822384426,
+      "grad_norm": 1.23190340238718,
+      "learning_rate": 8.177978349680727e-06,
+      "loss": 0.3644,
+      "step": 3107
+    },
+    {
+      "epoch": 0.3024817518248175,
+      "grad_norm": 1.4496645177592962,
+      "learning_rate": 8.176761365090503e-06,
+      "loss": 0.526,
+      "step": 3108
+    },
+    {
+      "epoch": 0.30257907542579077,
+      "grad_norm": 1.5209859048615393,
+      "learning_rate": 8.17554406482355e-06,
+      "loss": 0.3034,
+      "step": 3109
+    },
+    {
+      "epoch": 0.30267639902676396,
+      "grad_norm": 1.4404359772108442,
+      "learning_rate": 8.17432644900083e-06,
+      "loss": 0.4735,
+      "step": 3110
+    },
+    {
+      "epoch": 0.3027737226277372,
+      "grad_norm": 1.2693525922216498,
+      "learning_rate": 8.173108517743343e-06,
+      "loss": 0.4021,
+      "step": 3111
+    },
+    {
+      "epoch": 0.30287104622871047,
+      "grad_norm": 1.3995736051817393,
+      "learning_rate": 8.171890271172109e-06,
+      "loss": 0.3084,
+      "step": 3112
+    },
+    {
+      "epoch": 0.3029683698296837,
+      "grad_norm": 1.5690384436250255,
+      "learning_rate": 8.17067170940819e-06,
+      "loss": 0.4097,
+      "step": 3113
+    },
+    {
+      "epoch": 0.3030656934306569,
+      "grad_norm": 1.270566641334736,
+      "learning_rate": 8.169452832572676e-06,
+      "loss": 0.3813,
+      "step": 3114
+    },
+    {
+      "epoch": 0.3031630170316302,
+      "grad_norm": 1.1690990375599999,
+      "learning_rate": 8.168233640786682e-06,
+      "loss": 0.2898,
+      "step": 3115
+    },
+    {
+      "epoch": 0.30326034063260343,
+      "grad_norm": 1.5367454476066444,
+      "learning_rate": 8.167014134171367e-06,
+      "loss": 0.4167,
+      "step": 3116
+    },
+    {
+      "epoch": 0.30335766423357663,
+      "grad_norm": 1.113322849500334,
+      "learning_rate": 8.165794312847912e-06,
+      "loss": 0.3274,
+      "step": 3117
+    },
+    {
+      "epoch": 0.3034549878345499,
+      "grad_norm": 1.4711999953076527,
+      "learning_rate": 8.164574176937527e-06,
+      "loss": 0.368,
+      "step": 3118
+    },
+    {
+      "epoch": 0.30355231143552314,
+      "grad_norm": 1.4465621082621003,
+      "learning_rate": 8.163353726561462e-06,
+      "loss": 0.2719,
+      "step": 3119
+    },
+    {
+      "epoch": 0.30364963503649633,
+      "grad_norm": 1.5224694722189016,
+      "learning_rate": 8.162132961840994e-06,
+      "loss": 0.3296,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3037469586374696,
+      "grad_norm": 1.3713377819635104,
+      "learning_rate": 8.160911882897429e-06,
+      "loss": 0.3064,
+      "step": 3121
+    },
+    {
+      "epoch": 0.30384428223844284,
+      "grad_norm": 1.6461819951429466,
+      "learning_rate": 8.159690489852108e-06,
+      "loss": 0.3646,
+      "step": 3122
+    },
+    {
+      "epoch": 0.30394160583941604,
+      "grad_norm": 1.4328493269552467,
+      "learning_rate": 8.1584687828264e-06,
+      "loss": 0.4363,
+      "step": 3123
+    },
+    {
+      "epoch": 0.3040389294403893,
+      "grad_norm": 1.2811535124384867,
+      "learning_rate": 8.157246761941708e-06,
+      "loss": 0.4582,
+      "step": 3124
+    },
+    {
+      "epoch": 0.30413625304136255,
+      "grad_norm": 1.2144846492785035,
+      "learning_rate": 8.156024427319464e-06,
+      "loss": 0.2413,
+      "step": 3125
+    },
+    {
+      "epoch": 0.30423357664233575,
+      "grad_norm": 1.4112474441167293,
+      "learning_rate": 8.154801779081135e-06,
+      "loss": 0.4762,
+      "step": 3126
+    },
+    {
+      "epoch": 0.304330900243309,
+      "grad_norm": 1.4641495751020401,
+      "learning_rate": 8.153578817348213e-06,
+      "loss": 0.4905,
+      "step": 3127
+    },
+    {
+      "epoch": 0.30442822384428225,
+      "grad_norm": 1.7041758523831827,
+      "learning_rate": 8.152355542242226e-06,
+      "loss": 0.5396,
+      "step": 3128
+    },
+    {
+      "epoch": 0.30452554744525545,
+      "grad_norm": 1.2349793608481423,
+      "learning_rate": 8.151131953884728e-06,
+      "loss": 0.3847,
+      "step": 3129
+    },
+    {
+      "epoch": 0.3046228710462287,
+      "grad_norm": 1.4859954822671841,
+      "learning_rate": 8.149908052397314e-06,
+      "loss": 0.5907,
+      "step": 3130
+    },
+    {
+      "epoch": 0.30472019464720196,
+      "grad_norm": 1.2693166698162108,
+      "learning_rate": 8.148683837901599e-06,
+      "loss": 0.2636,
+      "step": 3131
+    },
+    {
+      "epoch": 0.30481751824817516,
+      "grad_norm": 1.1084842598244526,
+      "learning_rate": 8.147459310519238e-06,
+      "loss": 0.3103,
+      "step": 3132
+    },
+    {
+      "epoch": 0.3049148418491484,
+      "grad_norm": 1.2151450124002459,
+      "learning_rate": 8.146234470371908e-06,
+      "loss": 0.2734,
+      "step": 3133
+    },
+    {
+      "epoch": 0.30501216545012166,
+      "grad_norm": 1.3625938628176788,
+      "learning_rate": 8.145009317581328e-06,
+      "loss": 0.3757,
+      "step": 3134
+    },
+    {
+      "epoch": 0.3051094890510949,
+      "grad_norm": 1.3367991206467007,
+      "learning_rate": 8.143783852269239e-06,
+      "loss": 0.3469,
+      "step": 3135
+    },
+    {
+      "epoch": 0.3052068126520681,
+      "grad_norm": 1.5485997458565015,
+      "learning_rate": 8.142558074557413e-06,
+      "loss": 0.6068,
+      "step": 3136
+    },
+    {
+      "epoch": 0.30530413625304137,
+      "grad_norm": 1.4327175362669387,
+      "learning_rate": 8.141331984567661e-06,
+      "loss": 0.4495,
+      "step": 3137
+    },
+    {
+      "epoch": 0.3054014598540146,
+      "grad_norm": 1.4648525390361329,
+      "learning_rate": 8.140105582421819e-06,
+      "loss": 0.4855,
+      "step": 3138
+    },
+    {
+      "epoch": 0.3054987834549878,
+      "grad_norm": 1.147269868566856,
+      "learning_rate": 8.138878868241755e-06,
+      "loss": 0.3671,
+      "step": 3139
+    },
+    {
+      "epoch": 0.3055961070559611,
+      "grad_norm": 1.4274559741070532,
+      "learning_rate": 8.13765184214937e-06,
+      "loss": 0.4246,
+      "step": 3140
+    },
+    {
+      "epoch": 0.30569343065693433,
+      "grad_norm": 1.3864373579762495,
+      "learning_rate": 8.13642450426659e-06,
+      "loss": 0.5252,
+      "step": 3141
+    },
+    {
+      "epoch": 0.3057907542579075,
+      "grad_norm": 1.3508219371380046,
+      "learning_rate": 8.135196854715382e-06,
+      "loss": 0.4022,
+      "step": 3142
+    },
+    {
+      "epoch": 0.3058880778588808,
+      "grad_norm": 1.3974501891292628,
+      "learning_rate": 8.133968893617734e-06,
+      "loss": 0.4903,
+      "step": 3143
+    },
+    {
+      "epoch": 0.30598540145985403,
+      "grad_norm": 1.379672607479744,
+      "learning_rate": 8.132740621095672e-06,
+      "loss": 0.4389,
+      "step": 3144
+    },
+    {
+      "epoch": 0.30608272506082723,
+      "grad_norm": 1.5338640282858476,
+      "learning_rate": 8.131512037271248e-06,
+      "loss": 0.5719,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3061800486618005,
+      "grad_norm": 1.754315640729903,
+      "learning_rate": 8.130283142266549e-06,
+      "loss": 0.4684,
+      "step": 3146
+    },
+    {
+      "epoch": 0.30627737226277374,
+      "grad_norm": 1.3260485404955915,
+      "learning_rate": 8.129053936203688e-06,
+      "loss": 0.3967,
+      "step": 3147
+    },
+    {
+      "epoch": 0.30637469586374694,
+      "grad_norm": 1.3926987348333701,
+      "learning_rate": 8.127824419204818e-06,
+      "loss": 0.3916,
+      "step": 3148
+    },
+    {
+      "epoch": 0.3064720194647202,
+      "grad_norm": 1.4794110467900325,
+      "learning_rate": 8.126594591392108e-06,
+      "loss": 0.4127,
+      "step": 3149
+    },
+    {
+      "epoch": 0.30656934306569344,
+      "grad_norm": 1.5256531082933278,
+      "learning_rate": 8.125364452887775e-06,
+      "loss": 0.4219,
+      "step": 3150
+    },
+    {
+      "epoch": 0.30666666666666664,
+      "grad_norm": 1.2765687697220431,
+      "learning_rate": 8.124134003814054e-06,
+      "loss": 0.3482,
+      "step": 3151
+    },
+    {
+      "epoch": 0.3067639902676399,
+      "grad_norm": 1.345009147300955,
+      "learning_rate": 8.122903244293217e-06,
+      "loss": 0.2419,
+      "step": 3152
+    },
+    {
+      "epoch": 0.30686131386861315,
+      "grad_norm": 1.243655794496096,
+      "learning_rate": 8.121672174447566e-06,
+      "loss": 0.3132,
+      "step": 3153
+    },
+    {
+      "epoch": 0.30695863746958635,
+      "grad_norm": 1.4526276096090445,
+      "learning_rate": 8.120440794399432e-06,
+      "loss": 0.5369,
+      "step": 3154
+    },
+    {
+      "epoch": 0.3070559610705596,
+      "grad_norm": 1.4088986355103132,
+      "learning_rate": 8.119209104271177e-06,
+      "loss": 0.331,
+      "step": 3155
+    },
+    {
+      "epoch": 0.30715328467153286,
+      "grad_norm": 1.4915113265208189,
+      "learning_rate": 8.117977104185198e-06,
+      "loss": 0.6195,
+      "step": 3156
+    },
+    {
+      "epoch": 0.3072506082725061,
+      "grad_norm": 1.143472956889321,
+      "learning_rate": 8.116744794263916e-06,
+      "loss": 0.2632,
+      "step": 3157
+    },
+    {
+      "epoch": 0.3073479318734793,
+      "grad_norm": 1.3575606240914238,
+      "learning_rate": 8.11551217462979e-06,
+      "loss": 0.3927,
+      "step": 3158
+    },
+    {
+      "epoch": 0.30744525547445256,
+      "grad_norm": 1.2891794787417357,
+      "learning_rate": 8.114279245405301e-06,
+      "loss": 0.3766,
+      "step": 3159
+    },
+    {
+      "epoch": 0.3075425790754258,
+      "grad_norm": 1.1671813045475161,
+      "learning_rate": 8.113046006712973e-06,
+      "loss": 0.3527,
+      "step": 3160
+    },
+    {
+      "epoch": 0.307639902676399,
+      "grad_norm": 1.2346428689153102,
+      "learning_rate": 8.111812458675348e-06,
+      "loss": 0.456,
+      "step": 3161
+    },
+    {
+      "epoch": 0.30773722627737227,
+      "grad_norm": 1.543619139526521,
+      "learning_rate": 8.110578601415007e-06,
+      "loss": 0.419,
+      "step": 3162
+    },
+    {
+      "epoch": 0.3078345498783455,
+      "grad_norm": 1.5361457722305751,
+      "learning_rate": 8.109344435054557e-06,
+      "loss": 0.4477,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3079318734793187,
+      "grad_norm": 1.0948111699644958,
+      "learning_rate": 8.108109959716641e-06,
+      "loss": 0.3469,
+      "step": 3164
+    },
+    {
+      "epoch": 0.30802919708029197,
+      "grad_norm": 1.4900262169803653,
+      "learning_rate": 8.106875175523928e-06,
+      "loss": 0.5066,
+      "step": 3165
+    },
+    {
+      "epoch": 0.3081265206812652,
+      "grad_norm": 1.2143378476964652,
+      "learning_rate": 8.105640082599118e-06,
+      "loss": 0.4016,
+      "step": 3166
+    },
+    {
+      "epoch": 0.3082238442822384,
+      "grad_norm": 1.227404068812886,
+      "learning_rate": 8.104404681064943e-06,
+      "loss": 0.3408,
+      "step": 3167
+    },
+    {
+      "epoch": 0.3083211678832117,
+      "grad_norm": 1.273486832675327,
+      "learning_rate": 8.10316897104417e-06,
+      "loss": 0.3819,
+      "step": 3168
+    },
+    {
+      "epoch": 0.30841849148418493,
+      "grad_norm": 1.390509439874599,
+      "learning_rate": 8.101932952659586e-06,
+      "loss": 0.5108,
+      "step": 3169
+    },
+    {
+      "epoch": 0.30851581508515813,
+      "grad_norm": 1.1910701089910116,
+      "learning_rate": 8.100696626034019e-06,
+      "loss": 0.3579,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3086131386861314,
+      "grad_norm": 1.6328619260471173,
+      "learning_rate": 8.099459991290324e-06,
+      "loss": 0.666,
+      "step": 3171
+    },
+    {
+      "epoch": 0.30871046228710464,
+      "grad_norm": 1.4085180007277236,
+      "learning_rate": 8.09822304855138e-06,
+      "loss": 0.3684,
+      "step": 3172
+    },
+    {
+      "epoch": 0.30880778588807783,
+      "grad_norm": 1.5920530522626664,
+      "learning_rate": 8.096985797940111e-06,
+      "loss": 0.4499,
+      "step": 3173
+    },
+    {
+      "epoch": 0.3089051094890511,
+      "grad_norm": 1.2895313219583247,
+      "learning_rate": 8.09574823957946e-06,
+      "loss": 0.4939,
+      "step": 3174
+    },
+    {
+      "epoch": 0.30900243309002434,
+      "grad_norm": 1.5242111980147517,
+      "learning_rate": 8.094510373592403e-06,
+      "loss": 0.3223,
+      "step": 3175
+    },
+    {
+      "epoch": 0.30909975669099754,
+      "grad_norm": 1.3628460645839475,
+      "learning_rate": 8.093272200101946e-06,
+      "loss": 0.507,
+      "step": 3176
+    },
+    {
+      "epoch": 0.3091970802919708,
+      "grad_norm": 1.45716247785806,
+      "learning_rate": 8.092033719231134e-06,
+      "loss": 0.2011,
+      "step": 3177
+    },
+    {
+      "epoch": 0.30929440389294405,
+      "grad_norm": 1.0824220688323085,
+      "learning_rate": 8.090794931103026e-06,
+      "loss": 0.2127,
+      "step": 3178
+    },
+    {
+      "epoch": 0.3093917274939173,
+      "grad_norm": 1.3637285345889945,
+      "learning_rate": 8.089555835840728e-06,
+      "loss": 0.3567,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3094890510948905,
+      "grad_norm": 1.352681485055594,
+      "learning_rate": 8.088316433567369e-06,
+      "loss": 0.4403,
+      "step": 3180
+    },
+    {
+      "epoch": 0.30958637469586375,
+      "grad_norm": 1.5330943463849844,
+      "learning_rate": 8.087076724406106e-06,
+      "loss": 0.3379,
+      "step": 3181
+    },
+    {
+      "epoch": 0.309683698296837,
+      "grad_norm": 1.4102464472701088,
+      "learning_rate": 8.08583670848013e-06,
+      "loss": 0.5173,
+      "step": 3182
+    },
+    {
+      "epoch": 0.3097810218978102,
+      "grad_norm": 1.3268465957799758,
+      "learning_rate": 8.084596385912666e-06,
+      "loss": 0.2684,
+      "step": 3183
+    },
+    {
+      "epoch": 0.30987834549878346,
+      "grad_norm": 1.0612856713580272,
+      "learning_rate": 8.083355756826962e-06,
+      "loss": 0.2057,
+      "step": 3184
+    },
+    {
+      "epoch": 0.3099756690997567,
+      "grad_norm": 1.2705392445129486,
+      "learning_rate": 8.082114821346302e-06,
+      "loss": 0.4234,
+      "step": 3185
+    },
+    {
+      "epoch": 0.3100729927007299,
+      "grad_norm": 1.502946661443704,
+      "learning_rate": 8.080873579593997e-06,
+      "loss": 0.5134,
+      "step": 3186
+    },
+    {
+      "epoch": 0.31017031630170316,
+      "grad_norm": 1.4788874331253061,
+      "learning_rate": 8.079632031693392e-06,
+      "loss": 0.6157,
+      "step": 3187
+    },
+    {
+      "epoch": 0.3102676399026764,
+      "grad_norm": 1.0642136610663042,
+      "learning_rate": 8.078390177767858e-06,
+      "loss": 0.2667,
+      "step": 3188
+    },
+    {
+      "epoch": 0.3103649635036496,
+      "grad_norm": 1.2694637339318475,
+      "learning_rate": 8.0771480179408e-06,
+      "loss": 0.4189,
+      "step": 3189
+    },
+    {
+      "epoch": 0.31046228710462287,
+      "grad_norm": 1.40127369638598,
+      "learning_rate": 8.075905552335652e-06,
+      "loss": 0.6007,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3105596107055961,
+      "grad_norm": 1.4654952978073685,
+      "learning_rate": 8.07466278107588e-06,
+      "loss": 0.524,
+      "step": 3191
+    },
+    {
+      "epoch": 0.3106569343065693,
+      "grad_norm": 1.468706008576981,
+      "learning_rate": 8.073419704284977e-06,
+      "loss": 0.5511,
+      "step": 3192
+    },
+    {
+      "epoch": 0.3107542579075426,
+      "grad_norm": 1.3171427220237197,
+      "learning_rate": 8.072176322086468e-06,
+      "loss": 0.4903,
+      "step": 3193
+    },
+    {
+      "epoch": 0.31085158150851583,
+      "grad_norm": 1.5864389312753313,
+      "learning_rate": 8.07093263460391e-06,
+      "loss": 0.7036,
+      "step": 3194
+    },
+    {
+      "epoch": 0.310948905109489,
+      "grad_norm": 1.3892836370042843,
+      "learning_rate": 8.06968864196089e-06,
+      "loss": 0.4277,
+      "step": 3195
+    },
+    {
+      "epoch": 0.3110462287104623,
+      "grad_norm": 1.1676123850477602,
+      "learning_rate": 8.06844434428102e-06,
+      "loss": 0.2693,
+      "step": 3196
+    },
+    {
+      "epoch": 0.31114355231143553,
+      "grad_norm": 1.506360397408636,
+      "learning_rate": 8.067199741687951e-06,
+      "loss": 0.4425,
+      "step": 3197
+    },
+    {
+      "epoch": 0.31124087591240873,
+      "grad_norm": 1.9902872035508865,
+      "learning_rate": 8.065954834305359e-06,
+      "loss": 0.4464,
+      "step": 3198
+    },
+    {
+      "epoch": 0.311338199513382,
+      "grad_norm": 1.3738283887856941,
+      "learning_rate": 8.06470962225695e-06,
+      "loss": 0.2504,
+      "step": 3199
+    },
+    {
+      "epoch": 0.31143552311435524,
+      "grad_norm": 1.5173657046475728,
+      "learning_rate": 8.063464105666462e-06,
+      "loss": 0.4145,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3115328467153285,
+      "grad_norm": 1.5348915862029098,
+      "learning_rate": 8.062218284657663e-06,
+      "loss": 0.4182,
+      "step": 3201
+    },
+    {
+      "epoch": 0.3116301703163017,
+      "grad_norm": 1.6415893626851852,
+      "learning_rate": 8.06097215935435e-06,
+      "loss": 0.5928,
+      "step": 3202
+    },
+    {
+      "epoch": 0.31172749391727494,
+      "grad_norm": 1.2746939048596826,
+      "learning_rate": 8.059725729880354e-06,
+      "loss": 0.2945,
+      "step": 3203
+    },
+    {
+      "epoch": 0.3118248175182482,
+      "grad_norm": 1.354856369246625,
+      "learning_rate": 8.05847899635953e-06,
+      "loss": 0.445,
+      "step": 3204
+    },
+    {
+      "epoch": 0.3119221411192214,
+      "grad_norm": 1.3006695307047205,
+      "learning_rate": 8.057231958915767e-06,
+      "loss": 0.3558,
+      "step": 3205
+    },
+    {
+      "epoch": 0.31201946472019465,
+      "grad_norm": 1.294215405183638,
+      "learning_rate": 8.05598461767299e-06,
+      "loss": 0.3764,
+      "step": 3206
+    },
+    {
+      "epoch": 0.3121167883211679,
+      "grad_norm": 1.4713454344723826,
+      "learning_rate": 8.054736972755138e-06,
+      "loss": 0.4945,
+      "step": 3207
+    },
+    {
+      "epoch": 0.3122141119221411,
+      "grad_norm": 1.1708979284493786,
+      "learning_rate": 8.053489024286198e-06,
+      "loss": 0.2419,
+      "step": 3208
+    },
+    {
+      "epoch": 0.31231143552311436,
+      "grad_norm": 1.2933153684751388,
+      "learning_rate": 8.052240772390176e-06,
+      "loss": 0.4624,
+      "step": 3209
+    },
+    {
+      "epoch": 0.3124087591240876,
+      "grad_norm": 1.2538957446837722,
+      "learning_rate": 8.050992217191114e-06,
+      "loss": 0.3305,
+      "step": 3210
+    },
+    {
+      "epoch": 0.3125060827250608,
+      "grad_norm": 1.5362890980077515,
+      "learning_rate": 8.049743358813078e-06,
+      "loss": 0.5151,
+      "step": 3211
+    },
+    {
+      "epoch": 0.31260340632603406,
+      "grad_norm": 1.3913400504692353,
+      "learning_rate": 8.04849419738017e-06,
+      "loss": 0.355,
+      "step": 3212
+    },
+    {
+      "epoch": 0.3127007299270073,
+      "grad_norm": 1.3089021233990763,
+      "learning_rate": 8.04724473301652e-06,
+      "loss": 0.262,
+      "step": 3213
+    },
+    {
+      "epoch": 0.3127980535279805,
+      "grad_norm": 1.0542807465796966,
+      "learning_rate": 8.045994965846288e-06,
+      "loss": 0.3133,
+      "step": 3214
+    },
+    {
+      "epoch": 0.31289537712895377,
+      "grad_norm": 1.4749126602435412,
+      "learning_rate": 8.044744895993666e-06,
+      "loss": 0.46,
+      "step": 3215
+    },
+    {
+      "epoch": 0.312992700729927,
+      "grad_norm": 1.3746833003058674,
+      "learning_rate": 8.043494523582871e-06,
+      "loss": 0.427,
+      "step": 3216
+    },
+    {
+      "epoch": 0.3130900243309002,
+      "grad_norm": 1.3205796823520726,
+      "learning_rate": 8.042243848738153e-06,
+      "loss": 0.3354,
+      "step": 3217
+    },
+    {
+      "epoch": 0.31318734793187347,
+      "grad_norm": 1.2865057356076828,
+      "learning_rate": 8.040992871583797e-06,
+      "loss": 0.3941,
+      "step": 3218
+    },
+    {
+      "epoch": 0.3132846715328467,
+      "grad_norm": 1.1248365389432082,
+      "learning_rate": 8.039741592244108e-06,
+      "loss": 0.2628,
+      "step": 3219
+    },
+    {
+      "epoch": 0.3133819951338199,
+      "grad_norm": 1.4652698761705358,
+      "learning_rate": 8.03849001084343e-06,
+      "loss": 0.3564,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3134793187347932,
+      "grad_norm": 1.3953599364279132,
+      "learning_rate": 8.037238127506128e-06,
+      "loss": 0.4163,
+      "step": 3221
+    },
+    {
+      "epoch": 0.31357664233576643,
+      "grad_norm": 1.2419892638751415,
+      "learning_rate": 8.035985942356612e-06,
+      "loss": 0.354,
+      "step": 3222
+    },
+    {
+      "epoch": 0.3136739659367397,
+      "grad_norm": 1.715104485156596,
+      "learning_rate": 8.034733455519303e-06,
+      "loss": 0.2963,
+      "step": 3223
+    },
+    {
+      "epoch": 0.3137712895377129,
+      "grad_norm": 1.470040424076559,
+      "learning_rate": 8.033480667118667e-06,
+      "loss": 0.4648,
+      "step": 3224
+    },
+    {
+      "epoch": 0.31386861313868614,
+      "grad_norm": 1.4565317560800817,
+      "learning_rate": 8.032227577279191e-06,
+      "loss": 0.512,
+      "step": 3225
+    },
+    {
+      "epoch": 0.3139659367396594,
+      "grad_norm": 1.3742041452053566,
+      "learning_rate": 8.030974186125397e-06,
+      "loss": 0.3956,
+      "step": 3226
+    },
+    {
+      "epoch": 0.3140632603406326,
+      "grad_norm": 1.3788492769137004,
+      "learning_rate": 8.029720493781838e-06,
+      "loss": 0.4509,
+      "step": 3227
+    },
+    {
+      "epoch": 0.31416058394160584,
+      "grad_norm": 1.3858125546461868,
+      "learning_rate": 8.028466500373089e-06,
+      "loss": 0.2106,
+      "step": 3228
+    },
+    {
+      "epoch": 0.3142579075425791,
+      "grad_norm": 1.3840076863400683,
+      "learning_rate": 8.027212206023762e-06,
+      "loss": 0.3038,
+      "step": 3229
+    },
+    {
+      "epoch": 0.3143552311435523,
+      "grad_norm": 1.3152632009702614,
+      "learning_rate": 8.0259576108585e-06,
+      "loss": 0.4801,
+      "step": 3230
+    },
+    {
+      "epoch": 0.31445255474452555,
+      "grad_norm": 1.3788629368363385,
+      "learning_rate": 8.024702715001968e-06,
+      "loss": 0.4245,
+      "step": 3231
+    },
+    {
+      "epoch": 0.3145498783454988,
+      "grad_norm": 1.814337226771454,
+      "learning_rate": 8.023447518578868e-06,
+      "loss": 0.5632,
+      "step": 3232
+    },
+    {
+      "epoch": 0.314647201946472,
+      "grad_norm": 1.4232062511008752,
+      "learning_rate": 8.02219202171393e-06,
+      "loss": 0.3286,
+      "step": 3233
+    },
+    {
+      "epoch": 0.31474452554744525,
+      "grad_norm": 1.280903932308868,
+      "learning_rate": 8.020936224531912e-06,
+      "loss": 0.3626,
+      "step": 3234
+    },
+    {
+      "epoch": 0.3148418491484185,
+      "grad_norm": 1.0510702763807533,
+      "learning_rate": 8.019680127157607e-06,
+      "loss": 0.2524,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3149391727493917,
+      "grad_norm": 1.7478480050109064,
+      "learning_rate": 8.018423729715832e-06,
+      "loss": 0.4348,
+      "step": 3236
+    },
+    {
+      "epoch": 0.31503649635036496,
+      "grad_norm": 1.5359033994100462,
+      "learning_rate": 8.017167032331434e-06,
+      "loss": 0.4124,
+      "step": 3237
+    },
+    {
+      "epoch": 0.3151338199513382,
+      "grad_norm": 1.506288459888311,
+      "learning_rate": 8.015910035129294e-06,
+      "loss": 0.3261,
+      "step": 3238
+    },
+    {
+      "epoch": 0.3152311435523114,
+      "grad_norm": 1.2745124165082422,
+      "learning_rate": 8.01465273823432e-06,
+      "loss": 0.464,
+      "step": 3239
+    },
+    {
+      "epoch": 0.31532846715328466,
+      "grad_norm": 1.338919807585357,
+      "learning_rate": 8.01339514177145e-06,
+      "loss": 0.4172,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3154257907542579,
+      "grad_norm": 1.486780887173232,
+      "learning_rate": 8.012137245865654e-06,
+      "loss": 0.5408,
+      "step": 3241
+    },
+    {
+      "epoch": 0.31552311435523117,
+      "grad_norm": 1.3714950278620026,
+      "learning_rate": 8.010879050641927e-06,
+      "loss": 0.3436,
+      "step": 3242
+    },
+    {
+      "epoch": 0.31562043795620437,
+      "grad_norm": 1.631965892281063,
+      "learning_rate": 8.009620556225298e-06,
+      "loss": 0.4727,
+      "step": 3243
+    },
+    {
+      "epoch": 0.3157177615571776,
+      "grad_norm": 1.6149761911532763,
+      "learning_rate": 8.008361762740825e-06,
+      "loss": 0.4924,
+      "step": 3244
+    },
+    {
+      "epoch": 0.3158150851581509,
+      "grad_norm": 1.3425533377851226,
+      "learning_rate": 8.007102670313596e-06,
+      "loss": 0.3844,
+      "step": 3245
+    },
+    {
+      "epoch": 0.3159124087591241,
+      "grad_norm": 1.5650571116791179,
+      "learning_rate": 8.005843279068724e-06,
+      "loss": 0.5109,
+      "step": 3246
+    },
+    {
+      "epoch": 0.31600973236009733,
+      "grad_norm": 1.469317468630812,
+      "learning_rate": 8.004583589131359e-06,
+      "loss": 0.3981,
+      "step": 3247
+    },
+    {
+      "epoch": 0.3161070559610706,
+      "grad_norm": 1.3520956586482695,
+      "learning_rate": 8.003323600626675e-06,
+      "loss": 0.3628,
+      "step": 3248
+    },
+    {
+      "epoch": 0.3162043795620438,
+      "grad_norm": 1.306843877245721,
+      "learning_rate": 8.002063313679881e-06,
+      "loss": 0.3738,
+      "step": 3249
+    },
+    {
+      "epoch": 0.31630170316301703,
+      "grad_norm": 1.2062559137545288,
+      "learning_rate": 8.000802728416209e-06,
+      "loss": 0.3603,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3163990267639903,
+      "grad_norm": 1.2270301850514787,
+      "learning_rate": 7.999541844960926e-06,
+      "loss": 0.3444,
+      "step": 3251
+    },
+    {
+      "epoch": 0.3164963503649635,
+      "grad_norm": 1.6301989651574331,
+      "learning_rate": 7.998280663439325e-06,
+      "loss": 0.5442,
+      "step": 3252
+    },
+    {
+      "epoch": 0.31659367396593674,
+      "grad_norm": 1.3633091002846736,
+      "learning_rate": 7.997019183976732e-06,
+      "loss": 0.4596,
+      "step": 3253
+    },
+    {
+      "epoch": 0.31669099756691,
+      "grad_norm": 1.0930243128300028,
+      "learning_rate": 7.9957574066985e-06,
+      "loss": 0.259,
+      "step": 3254
+    },
+    {
+      "epoch": 0.3167883211678832,
+      "grad_norm": 1.3445912322829077,
+      "learning_rate": 7.994495331730014e-06,
+      "loss": 0.438,
+      "step": 3255
+    },
+    {
+      "epoch": 0.31688564476885644,
+      "grad_norm": 1.3651374487790005,
+      "learning_rate": 7.993232959196687e-06,
+      "loss": 0.4589,
+      "step": 3256
+    },
+    {
+      "epoch": 0.3169829683698297,
+      "grad_norm": 1.3424961538987858,
+      "learning_rate": 7.99197028922396e-06,
+      "loss": 0.4367,
+      "step": 3257
+    },
+    {
+      "epoch": 0.3170802919708029,
+      "grad_norm": 1.5932344837064665,
+      "learning_rate": 7.990707321937308e-06,
+      "loss": 0.6921,
+      "step": 3258
+    },
+    {
+      "epoch": 0.31717761557177615,
+      "grad_norm": 1.475521709829228,
+      "learning_rate": 7.989444057462228e-06,
+      "loss": 0.4759,
+      "step": 3259
+    },
+    {
+      "epoch": 0.3172749391727494,
+      "grad_norm": 1.2971398875872913,
+      "learning_rate": 7.988180495924256e-06,
+      "loss": 0.4588,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3173722627737226,
+      "grad_norm": 1.4756687426354647,
+      "learning_rate": 7.986916637448953e-06,
+      "loss": 0.4776,
+      "step": 3261
+    },
+    {
+      "epoch": 0.31746958637469586,
+      "grad_norm": 1.6271934377592354,
+      "learning_rate": 7.985652482161907e-06,
+      "loss": 0.4979,
+      "step": 3262
+    },
+    {
+      "epoch": 0.3175669099756691,
+      "grad_norm": 1.6597644298654206,
+      "learning_rate": 7.984388030188739e-06,
+      "loss": 0.6091,
+      "step": 3263
+    },
+    {
+      "epoch": 0.31766423357664236,
+      "grad_norm": 1.3383482658500123,
+      "learning_rate": 7.983123281655097e-06,
+      "loss": 0.4371,
+      "step": 3264
+    },
+    {
+      "epoch": 0.31776155717761556,
+      "grad_norm": 1.4591756386356707,
+      "learning_rate": 7.981858236686661e-06,
+      "loss": 0.4888,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3178588807785888,
+      "grad_norm": 1.1677482677553854,
+      "learning_rate": 7.98059289540914e-06,
+      "loss": 0.348,
+      "step": 3266
+    },
+    {
+      "epoch": 0.31795620437956207,
+      "grad_norm": 1.280054884121347,
+      "learning_rate": 7.97932725794827e-06,
+      "loss": 0.3909,
+      "step": 3267
+    },
+    {
+      "epoch": 0.31805352798053527,
+      "grad_norm": 1.2818245878680554,
+      "learning_rate": 7.97806132442982e-06,
+      "loss": 0.3432,
+      "step": 3268
+    },
+    {
+      "epoch": 0.3181508515815085,
+      "grad_norm": 1.2511980990717368,
+      "learning_rate": 7.976795094979586e-06,
+      "loss": 0.398,
+      "step": 3269
+    },
+    {
+      "epoch": 0.3182481751824818,
+      "grad_norm": 1.1398243641659185,
+      "learning_rate": 7.975528569723391e-06,
+      "loss": 0.3561,
+      "step": 3270
+    },
+    {
+      "epoch": 0.31834549878345497,
+      "grad_norm": 1.4375913010503336,
+      "learning_rate": 7.974261748787096e-06,
+      "loss": 0.4341,
+      "step": 3271
+    },
+    {
+      "epoch": 0.3184428223844282,
+      "grad_norm": 1.5232808350435216,
+      "learning_rate": 7.972994632296583e-06,
+      "loss": 0.443,
+      "step": 3272
+    },
+    {
+      "epoch": 0.3185401459854015,
+      "grad_norm": 1.1586214953526035,
+      "learning_rate": 7.971727220377765e-06,
+      "loss": 0.3709,
+      "step": 3273
+    },
+    {
+      "epoch": 0.3186374695863747,
+      "grad_norm": 1.5274740880588324,
+      "learning_rate": 7.970459513156587e-06,
+      "loss": 0.3699,
+      "step": 3274
+    },
+    {
+      "epoch": 0.31873479318734793,
+      "grad_norm": 1.3981699767571285,
+      "learning_rate": 7.969191510759021e-06,
+      "loss": 0.3678,
+      "step": 3275
+    },
+    {
+      "epoch": 0.3188321167883212,
+      "grad_norm": 1.934723248663663,
+      "learning_rate": 7.96792321331107e-06,
+      "loss": 0.5528,
+      "step": 3276
+    },
+    {
+      "epoch": 0.3189294403892944,
+      "grad_norm": 1.3063523868424662,
+      "learning_rate": 7.966654620938765e-06,
+      "loss": 0.381,
+      "step": 3277
+    },
+    {
+      "epoch": 0.31902676399026764,
+      "grad_norm": 1.4316009890701873,
+      "learning_rate": 7.965385733768166e-06,
+      "loss": 0.3462,
+      "step": 3278
+    },
+    {
+      "epoch": 0.3191240875912409,
+      "grad_norm": 1.3925079285209303,
+      "learning_rate": 7.964116551925365e-06,
+      "loss": 0.3468,
+      "step": 3279
+    },
+    {
+      "epoch": 0.3192214111922141,
+      "grad_norm": 1.5929020888752208,
+      "learning_rate": 7.96284707553648e-06,
+      "loss": 0.5416,
+      "step": 3280
+    },
+    {
+      "epoch": 0.31931873479318734,
+      "grad_norm": 1.3866900373898063,
+      "learning_rate": 7.961577304727659e-06,
+      "loss": 0.3982,
+      "step": 3281
+    },
+    {
+      "epoch": 0.3194160583941606,
+      "grad_norm": 1.4962340605356037,
+      "learning_rate": 7.960307239625082e-06,
+      "loss": 0.4023,
+      "step": 3282
+    },
+    {
+      "epoch": 0.3195133819951338,
+      "grad_norm": 1.4057559523045156,
+      "learning_rate": 7.959036880354955e-06,
+      "loss": 0.495,
+      "step": 3283
+    },
+    {
+      "epoch": 0.31961070559610705,
+      "grad_norm": 1.2622065590243314,
+      "learning_rate": 7.957766227043514e-06,
+      "loss": 0.3581,
+      "step": 3284
+    },
+    {
+      "epoch": 0.3197080291970803,
+      "grad_norm": 1.5090453488845967,
+      "learning_rate": 7.956495279817026e-06,
+      "loss": 0.455,
+      "step": 3285
+    },
+    {
+      "epoch": 0.31980535279805355,
+      "grad_norm": 1.3133879337401893,
+      "learning_rate": 7.955224038801785e-06,
+      "loss": 0.4625,
+      "step": 3286
+    },
+    {
+      "epoch": 0.31990267639902675,
+      "grad_norm": 2.521541296377749,
+      "learning_rate": 7.953952504124114e-06,
+      "loss": 0.4415,
+      "step": 3287
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3567937262050411,
+      "learning_rate": 7.952680675910365e-06,
+      "loss": 0.3309,
+      "step": 3288
+    },
+    {
+      "epoch": 0.32009732360097326,
+      "grad_norm": 1.5421944908903493,
+      "learning_rate": 7.951408554286926e-06,
+      "loss": 0.4589,
+      "step": 3289
+    },
+    {
+      "epoch": 0.32019464720194646,
+      "grad_norm": 1.5998274173642424,
+      "learning_rate": 7.950136139380204e-06,
+      "loss": 0.5359,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3202919708029197,
+      "grad_norm": 1.2725707304657317,
+      "learning_rate": 7.948863431316639e-06,
+      "loss": 0.3625,
+      "step": 3291
+    },
+    {
+      "epoch": 0.32038929440389297,
+      "grad_norm": 1.4290851095226622,
+      "learning_rate": 7.947590430222702e-06,
+      "loss": 0.4872,
+      "step": 3292
+    },
+    {
+      "epoch": 0.32048661800486616,
+      "grad_norm": 1.3420498316619087,
+      "learning_rate": 7.946317136224894e-06,
+      "loss": 0.2389,
+      "step": 3293
+    },
+    {
+      "epoch": 0.3205839416058394,
+      "grad_norm": 1.5002712163507215,
+      "learning_rate": 7.94504354944974e-06,
+      "loss": 0.5157,
+      "step": 3294
+    },
+    {
+      "epoch": 0.32068126520681267,
+      "grad_norm": 1.660535414536618,
+      "learning_rate": 7.9437696700238e-06,
+      "loss": 0.4267,
+      "step": 3295
+    },
+    {
+      "epoch": 0.32077858880778587,
+      "grad_norm": 1.2370297819791147,
+      "learning_rate": 7.942495498073657e-06,
+      "loss": 0.3355,
+      "step": 3296
+    },
+    {
+      "epoch": 0.3208759124087591,
+      "grad_norm": 1.1415276287275913,
+      "learning_rate": 7.941221033725928e-06,
+      "loss": 0.2944,
+      "step": 3297
+    },
+    {
+      "epoch": 0.3209732360097324,
+      "grad_norm": 1.5554788387015477,
+      "learning_rate": 7.939946277107258e-06,
+      "loss": 0.4871,
+      "step": 3298
+    },
+    {
+      "epoch": 0.3210705596107056,
+      "grad_norm": 1.2985241068062738,
+      "learning_rate": 7.938671228344319e-06,
+      "loss": 0.3143,
+      "step": 3299
+    },
+    {
+      "epoch": 0.32116788321167883,
+      "grad_norm": 1.301901578297674,
+      "learning_rate": 7.937395887563812e-06,
+      "loss": 0.3965,
+      "step": 3300
+    },
+    {
+      "epoch": 0.3212652068126521,
+      "grad_norm": 0.9529902878913864,
+      "learning_rate": 7.936120254892471e-06,
+      "loss": 0.3083,
+      "step": 3301
+    },
+    {
+      "epoch": 0.3213625304136253,
+      "grad_norm": 1.442015067028423,
+      "learning_rate": 7.934844330457056e-06,
+      "loss": 0.4318,
+      "step": 3302
+    },
+    {
+      "epoch": 0.32145985401459853,
+      "grad_norm": 0.8997835110854661,
+      "learning_rate": 7.933568114384358e-06,
+      "loss": 0.2885,
+      "step": 3303
+    },
+    {
+      "epoch": 0.3215571776155718,
+      "grad_norm": 1.5293826180265608,
+      "learning_rate": 7.932291606801192e-06,
+      "loss": 0.5437,
+      "step": 3304
+    },
+    {
+      "epoch": 0.321654501216545,
+      "grad_norm": 1.2264156375116992,
+      "learning_rate": 7.931014807834405e-06,
+      "loss": 0.4001,
+      "step": 3305
+    },
+    {
+      "epoch": 0.32175182481751824,
+      "grad_norm": 1.260350527748902,
+      "learning_rate": 7.929737717610878e-06,
+      "loss": 0.3847,
+      "step": 3306
+    },
+    {
+      "epoch": 0.3218491484184915,
+      "grad_norm": 1.1346472253273232,
+      "learning_rate": 7.92846033625751e-06,
+      "loss": 0.3766,
+      "step": 3307
+    },
+    {
+      "epoch": 0.32194647201946475,
+      "grad_norm": 1.1035142206503785,
+      "learning_rate": 7.927182663901241e-06,
+      "loss": 0.369,
+      "step": 3308
+    },
+    {
+      "epoch": 0.32204379562043794,
+      "grad_norm": 1.2980599562576733,
+      "learning_rate": 7.92590470066903e-06,
+      "loss": 0.3982,
+      "step": 3309
+    },
+    {
+      "epoch": 0.3221411192214112,
+      "grad_norm": 1.1742301163722888,
+      "learning_rate": 7.924626446687871e-06,
+      "loss": 0.3423,
+      "step": 3310
+    },
+    {
+      "epoch": 0.32223844282238445,
+      "grad_norm": 1.8345279558348917,
+      "learning_rate": 7.923347902084784e-06,
+      "loss": 0.3145,
+      "step": 3311
+    },
+    {
+      "epoch": 0.32233576642335765,
+      "grad_norm": 1.562283157560816,
+      "learning_rate": 7.92206906698682e-06,
+      "loss": 0.4308,
+      "step": 3312
+    },
+    {
+      "epoch": 0.3224330900243309,
+      "grad_norm": 1.568673716358362,
+      "learning_rate": 7.920789941521053e-06,
+      "loss": 0.7025,
+      "step": 3313
+    },
+    {
+      "epoch": 0.32253041362530416,
+      "grad_norm": 1.409203500738135,
+      "learning_rate": 7.9195105258146e-06,
+      "loss": 0.459,
+      "step": 3314
+    },
+    {
+      "epoch": 0.32262773722627736,
+      "grad_norm": 0.880538645206563,
+      "learning_rate": 7.918230819994589e-06,
+      "loss": 0.2786,
+      "step": 3315
+    },
+    {
+      "epoch": 0.3227250608272506,
+      "grad_norm": 1.1906838149715093,
+      "learning_rate": 7.916950824188188e-06,
+      "loss": 0.2686,
+      "step": 3316
+    },
+    {
+      "epoch": 0.32282238442822386,
+      "grad_norm": 1.5111358859487132,
+      "learning_rate": 7.91567053852259e-06,
+      "loss": 0.5147,
+      "step": 3317
+    },
+    {
+      "epoch": 0.32291970802919706,
+      "grad_norm": 1.1464259378074475,
+      "learning_rate": 7.914389963125018e-06,
+      "loss": 0.2685,
+      "step": 3318
+    },
+    {
+      "epoch": 0.3230170316301703,
+      "grad_norm": 1.581506486679798,
+      "learning_rate": 7.913109098122726e-06,
+      "loss": 0.5854,
+      "step": 3319
+    },
+    {
+      "epoch": 0.32311435523114357,
+      "grad_norm": 1.3242645255760028,
+      "learning_rate": 7.91182794364299e-06,
+      "loss": 0.2315,
+      "step": 3320
+    },
+    {
+      "epoch": 0.32321167883211677,
+      "grad_norm": 1.6094300344001513,
+      "learning_rate": 7.910546499813125e-06,
+      "loss": 0.4739,
+      "step": 3321
+    },
+    {
+      "epoch": 0.32330900243309,
+      "grad_norm": 1.3446882210600364,
+      "learning_rate": 7.909264766760462e-06,
+      "loss": 0.4145,
+      "step": 3322
+    },
+    {
+      "epoch": 0.3234063260340633,
+      "grad_norm": 1.282785844235514,
+      "learning_rate": 7.907982744612373e-06,
+      "loss": 0.4324,
+      "step": 3323
+    },
+    {
+      "epoch": 0.32350364963503647,
+      "grad_norm": 1.8199182173411141,
+      "learning_rate": 7.90670043349625e-06,
+      "loss": 0.2738,
+      "step": 3324
+    },
+    {
+      "epoch": 0.3236009732360097,
+      "grad_norm": 1.3779170997525898,
+      "learning_rate": 7.90541783353952e-06,
+      "loss": 0.4386,
+      "step": 3325
+    },
+    {
+      "epoch": 0.323698296836983,
+      "grad_norm": 1.4651475108226641,
+      "learning_rate": 7.904134944869631e-06,
+      "loss": 0.2272,
+      "step": 3326
+    },
+    {
+      "epoch": 0.3237956204379562,
+      "grad_norm": 1.4444649350514327,
+      "learning_rate": 7.902851767614069e-06,
+      "loss": 0.3631,
+      "step": 3327
+    },
+    {
+      "epoch": 0.32389294403892943,
+      "grad_norm": 1.6625856822026597,
+      "learning_rate": 7.901568301900343e-06,
+      "loss": 0.3649,
+      "step": 3328
+    },
+    {
+      "epoch": 0.3239902676399027,
+      "grad_norm": 1.4506728288423798,
+      "learning_rate": 7.900284547855992e-06,
+      "loss": 0.3231,
+      "step": 3329
+    },
+    {
+      "epoch": 0.32408759124087594,
+      "grad_norm": 1.5932450335999997,
+      "learning_rate": 7.899000505608583e-06,
+      "loss": 0.6145,
+      "step": 3330
+    },
+    {
+      "epoch": 0.32418491484184914,
+      "grad_norm": 1.466657731000094,
+      "learning_rate": 7.89771617528571e-06,
+      "loss": 0.4498,
+      "step": 3331
+    },
+    {
+      "epoch": 0.3242822384428224,
+      "grad_norm": 1.1444340687397683,
+      "learning_rate": 7.896431557015001e-06,
+      "loss": 0.3953,
+      "step": 3332
+    },
+    {
+      "epoch": 0.32437956204379564,
+      "grad_norm": 1.2696327521021862,
+      "learning_rate": 7.895146650924106e-06,
+      "loss": 0.3974,
+      "step": 3333
+    },
+    {
+      "epoch": 0.32447688564476884,
+      "grad_norm": 1.0672959190242737,
+      "learning_rate": 7.893861457140711e-06,
+      "loss": 0.3147,
+      "step": 3334
+    },
+    {
+      "epoch": 0.3245742092457421,
+      "grad_norm": 1.5878275615936528,
+      "learning_rate": 7.892575975792524e-06,
+      "loss": 0.5637,
+      "step": 3335
+    },
+    {
+      "epoch": 0.32467153284671535,
+      "grad_norm": 1.3553067860868797,
+      "learning_rate": 7.891290207007284e-06,
+      "loss": 0.3979,
+      "step": 3336
+    },
+    {
+      "epoch": 0.32476885644768855,
+      "grad_norm": 1.3585205023281657,
+      "learning_rate": 7.890004150912758e-06,
+      "loss": 0.5408,
+      "step": 3337
+    },
+    {
+      "epoch": 0.3248661800486618,
+      "grad_norm": 1.3335333713837063,
+      "learning_rate": 7.888717807636745e-06,
+      "loss": 0.5097,
+      "step": 3338
+    },
+    {
+      "epoch": 0.32496350364963505,
+      "grad_norm": 1.6084982696846433,
+      "learning_rate": 7.887431177307067e-06,
+      "loss": 0.6652,
+      "step": 3339
+    },
+    {
+      "epoch": 0.32506082725060825,
+      "grad_norm": 1.5254713518221517,
+      "learning_rate": 7.886144260051577e-06,
+      "loss": 0.5413,
+      "step": 3340
+    },
+    {
+      "epoch": 0.3251581508515815,
+      "grad_norm": 1.2319691547427678,
+      "learning_rate": 7.88485705599816e-06,
+      "loss": 0.3669,
+      "step": 3341
+    },
+    {
+      "epoch": 0.32525547445255476,
+      "grad_norm": 1.8060436355287317,
+      "learning_rate": 7.883569565274722e-06,
+      "loss": 0.332,
+      "step": 3342
+    },
+    {
+      "epoch": 0.32535279805352796,
+      "grad_norm": 1.2724687132520958,
+      "learning_rate": 7.882281788009207e-06,
+      "loss": 0.4156,
+      "step": 3343
+    },
+    {
+      "epoch": 0.3254501216545012,
+      "grad_norm": 1.2678653056689784,
+      "learning_rate": 7.880993724329578e-06,
+      "loss": 0.34,
+      "step": 3344
+    },
+    {
+      "epoch": 0.32554744525547447,
+      "grad_norm": 1.6551598614012555,
+      "learning_rate": 7.879705374363831e-06,
+      "loss": 0.4642,
+      "step": 3345
+    },
+    {
+      "epoch": 0.32564476885644766,
+      "grad_norm": 1.3206920191183078,
+      "learning_rate": 7.878416738239991e-06,
+      "loss": 0.3755,
+      "step": 3346
+    },
+    {
+      "epoch": 0.3257420924574209,
+      "grad_norm": 1.3858157374277495,
+      "learning_rate": 7.877127816086109e-06,
+      "loss": 0.3394,
+      "step": 3347
+    },
+    {
+      "epoch": 0.32583941605839417,
+      "grad_norm": 2.0774474222482286,
+      "learning_rate": 7.87583860803027e-06,
+      "loss": 0.4237,
+      "step": 3348
+    },
+    {
+      "epoch": 0.32593673965936737,
+      "grad_norm": 2.373250216339497,
+      "learning_rate": 7.87454911420058e-06,
+      "loss": 0.4854,
+      "step": 3349
+    },
+    {
+      "epoch": 0.3260340632603406,
+      "grad_norm": 1.2888779016735306,
+      "learning_rate": 7.873259334725177e-06,
+      "loss": 0.2953,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3261313868613139,
+      "grad_norm": 1.4655511800886896,
+      "learning_rate": 7.87196926973223e-06,
+      "loss": 0.5252,
+      "step": 3351
+    },
+    {
+      "epoch": 0.32622871046228713,
+      "grad_norm": 1.3179348636560488,
+      "learning_rate": 7.870678919349929e-06,
+      "loss": 0.3587,
+      "step": 3352
+    },
+    {
+      "epoch": 0.32632603406326033,
+      "grad_norm": 1.5442264887171864,
+      "learning_rate": 7.869388283706501e-06,
+      "loss": 0.3808,
+      "step": 3353
+    },
+    {
+      "epoch": 0.3264233576642336,
+      "grad_norm": 0.909176934030135,
+      "learning_rate": 7.868097362930194e-06,
+      "loss": 0.1721,
+      "step": 3354
+    },
+    {
+      "epoch": 0.32652068126520684,
+      "grad_norm": 1.5955557971117078,
+      "learning_rate": 7.866806157149291e-06,
+      "loss": 0.5127,
+      "step": 3355
+    },
+    {
+      "epoch": 0.32661800486618003,
+      "grad_norm": 1.260336198155953,
+      "learning_rate": 7.865514666492096e-06,
+      "loss": 0.2699,
+      "step": 3356
+    },
+    {
+      "epoch": 0.3267153284671533,
+      "grad_norm": 1.5004384671553455,
+      "learning_rate": 7.864222891086948e-06,
+      "loss": 0.3168,
+      "step": 3357
+    },
+    {
+      "epoch": 0.32681265206812654,
+      "grad_norm": 1.261076205541224,
+      "learning_rate": 7.862930831062211e-06,
+      "loss": 0.3678,
+      "step": 3358
+    },
+    {
+      "epoch": 0.32690997566909974,
+      "grad_norm": 1.637405579152474,
+      "learning_rate": 7.861638486546279e-06,
+      "loss": 0.4613,
+      "step": 3359
+    },
+    {
+      "epoch": 0.327007299270073,
+      "grad_norm": 1.3584995739472485,
+      "learning_rate": 7.860345857667571e-06,
+      "loss": 0.3229,
+      "step": 3360
+    },
+    {
+      "epoch": 0.32710462287104625,
+      "grad_norm": 1.5428951197572305,
+      "learning_rate": 7.859052944554537e-06,
+      "loss": 0.4904,
+      "step": 3361
+    },
+    {
+      "epoch": 0.32720194647201944,
+      "grad_norm": 1.2674356859582525,
+      "learning_rate": 7.857759747335652e-06,
+      "loss": 0.2942,
+      "step": 3362
+    },
+    {
+      "epoch": 0.3272992700729927,
+      "grad_norm": 1.5802075836374327,
+      "learning_rate": 7.856466266139426e-06,
+      "loss": 0.2949,
+      "step": 3363
+    },
+    {
+      "epoch": 0.32739659367396595,
+      "grad_norm": 1.8209546966028776,
+      "learning_rate": 7.855172501094394e-06,
+      "loss": 0.5036,
+      "step": 3364
+    },
+    {
+      "epoch": 0.32749391727493915,
+      "grad_norm": 1.5829487718400805,
+      "learning_rate": 7.853878452329113e-06,
+      "loss": 0.3638,
+      "step": 3365
+    },
+    {
+      "epoch": 0.3275912408759124,
+      "grad_norm": 1.439883288183698,
+      "learning_rate": 7.852584119972178e-06,
+      "loss": 0.4529,
+      "step": 3366
+    },
+    {
+      "epoch": 0.32768856447688566,
+      "grad_norm": 1.188441154560626,
+      "learning_rate": 7.851289504152201e-06,
+      "loss": 0.1984,
+      "step": 3367
+    },
+    {
+      "epoch": 0.32778588807785886,
+      "grad_norm": 1.3841398947965158,
+      "learning_rate": 7.84999460499784e-06,
+      "loss": 0.3274,
+      "step": 3368
+    },
+    {
+      "epoch": 0.3278832116788321,
+      "grad_norm": 1.3784180119913427,
+      "learning_rate": 7.848699422637757e-06,
+      "loss": 0.5186,
+      "step": 3369
+    },
+    {
+      "epoch": 0.32798053527980536,
+      "grad_norm": 1.6167829071938324,
+      "learning_rate": 7.847403957200667e-06,
+      "loss": 0.5905,
+      "step": 3370
+    },
+    {
+      "epoch": 0.32807785888077856,
+      "grad_norm": 1.5035664756701093,
+      "learning_rate": 7.846108208815292e-06,
+      "loss": 0.3502,
+      "step": 3371
+    },
+    {
+      "epoch": 0.3281751824817518,
+      "grad_norm": 1.3836098651173667,
+      "learning_rate": 7.844812177610398e-06,
+      "loss": 0.426,
+      "step": 3372
+    },
+    {
+      "epoch": 0.32827250608272507,
+      "grad_norm": 1.33208856990685,
+      "learning_rate": 7.843515863714766e-06,
+      "loss": 0.38,
+      "step": 3373
+    },
+    {
+      "epoch": 0.3283698296836983,
+      "grad_norm": 1.3222208170433662,
+      "learning_rate": 7.842219267257216e-06,
+      "loss": 0.33,
+      "step": 3374
+    },
+    {
+      "epoch": 0.3284671532846715,
+      "grad_norm": 1.26421550866771,
+      "learning_rate": 7.84092238836659e-06,
+      "loss": 0.3682,
+      "step": 3375
+    },
+    {
+      "epoch": 0.3285644768856448,
+      "grad_norm": 1.3594451535417975,
+      "learning_rate": 7.839625227171762e-06,
+      "loss": 0.4504,
+      "step": 3376
+    },
+    {
+      "epoch": 0.328661800486618,
+      "grad_norm": 1.2038996790979526,
+      "learning_rate": 7.838327783801627e-06,
+      "loss": 0.3675,
+      "step": 3377
+    },
+    {
+      "epoch": 0.3287591240875912,
+      "grad_norm": 1.2523761100181583,
+      "learning_rate": 7.837030058385117e-06,
+      "loss": 0.2582,
+      "step": 3378
+    },
+    {
+      "epoch": 0.3288564476885645,
+      "grad_norm": 1.38623949822592,
+      "learning_rate": 7.835732051051188e-06,
+      "loss": 0.426,
+      "step": 3379
+    },
+    {
+      "epoch": 0.32895377128953773,
+      "grad_norm": 1.6752196039887233,
+      "learning_rate": 7.834433761928819e-06,
+      "loss": 0.5995,
+      "step": 3380
+    },
+    {
+      "epoch": 0.32905109489051093,
+      "grad_norm": 1.4769893614554368,
+      "learning_rate": 7.833135191147027e-06,
+      "loss": 0.4434,
+      "step": 3381
+    },
+    {
+      "epoch": 0.3291484184914842,
+      "grad_norm": 1.3473793961690093,
+      "learning_rate": 7.831836338834851e-06,
+      "loss": 0.4064,
+      "step": 3382
+    },
+    {
+      "epoch": 0.32924574209245744,
+      "grad_norm": 1.4564612861566626,
+      "learning_rate": 7.830537205121354e-06,
+      "loss": 0.5275,
+      "step": 3383
+    },
+    {
+      "epoch": 0.32934306569343064,
+      "grad_norm": 1.4662477809046923,
+      "learning_rate": 7.829237790135638e-06,
+      "loss": 0.3745,
+      "step": 3384
+    },
+    {
+      "epoch": 0.3294403892944039,
+      "grad_norm": 1.3883338656353856,
+      "learning_rate": 7.827938094006822e-06,
+      "loss": 0.4361,
+      "step": 3385
+    },
+    {
+      "epoch": 0.32953771289537714,
+      "grad_norm": 1.2360062745222065,
+      "learning_rate": 7.826638116864061e-06,
+      "loss": 0.2936,
+      "step": 3386
+    },
+    {
+      "epoch": 0.32963503649635034,
+      "grad_norm": 1.2636393287865908,
+      "learning_rate": 7.82533785883653e-06,
+      "loss": 0.3816,
+      "step": 3387
+    },
+    {
+      "epoch": 0.3297323600973236,
+      "grad_norm": 1.5704874728825693,
+      "learning_rate": 7.824037320053442e-06,
+      "loss": 0.4946,
+      "step": 3388
+    },
+    {
+      "epoch": 0.32982968369829685,
+      "grad_norm": 1.5450366878165769,
+      "learning_rate": 7.822736500644028e-06,
+      "loss": 0.5973,
+      "step": 3389
+    },
+    {
+      "epoch": 0.32992700729927005,
+      "grad_norm": 1.570140468606553,
+      "learning_rate": 7.821435400737555e-06,
+      "loss": 0.6187,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3300243309002433,
+      "grad_norm": 1.404973531589098,
+      "learning_rate": 7.820134020463311e-06,
+      "loss": 0.4404,
+      "step": 3391
+    },
+    {
+      "epoch": 0.33012165450121655,
+      "grad_norm": 1.4221866811013593,
+      "learning_rate": 7.818832359950615e-06,
+      "loss": 0.4375,
+      "step": 3392
+    },
+    {
+      "epoch": 0.3302189781021898,
+      "grad_norm": 1.3514761483715907,
+      "learning_rate": 7.817530419328815e-06,
+      "loss": 0.4633,
+      "step": 3393
+    },
+    {
+      "epoch": 0.330316301703163,
+      "grad_norm": 1.4122938191319212,
+      "learning_rate": 7.816228198727287e-06,
+      "loss": 0.4735,
+      "step": 3394
+    },
+    {
+      "epoch": 0.33041362530413626,
+      "grad_norm": 1.2807472917541904,
+      "learning_rate": 7.814925698275432e-06,
+      "loss": 0.2993,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3305109489051095,
+      "grad_norm": 1.2374164437493267,
+      "learning_rate": 7.813622918102679e-06,
+      "loss": 0.4486,
+      "step": 3396
+    },
+    {
+      "epoch": 0.3306082725060827,
+      "grad_norm": 1.4923726710921128,
+      "learning_rate": 7.812319858338486e-06,
+      "loss": 0.3976,
+      "step": 3397
+    },
+    {
+      "epoch": 0.33070559610705597,
+      "grad_norm": 1.4652422146853137,
+      "learning_rate": 7.811016519112342e-06,
+      "loss": 0.509,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3308029197080292,
+      "grad_norm": 1.3523346564010856,
+      "learning_rate": 7.80971290055376e-06,
+      "loss": 0.4045,
+      "step": 3399
+    },
+    {
+      "epoch": 0.3309002433090024,
+      "grad_norm": 1.4034359644709637,
+      "learning_rate": 7.808409002792277e-06,
+      "loss": 0.5252,
+      "step": 3400
+    },
+    {
+      "epoch": 0.33099756690997567,
+      "grad_norm": 1.5977042267924388,
+      "learning_rate": 7.807104825957466e-06,
+      "loss": 0.5708,
+      "step": 3401
+    },
+    {
+      "epoch": 0.3310948905109489,
+      "grad_norm": 1.194169039851455,
+      "learning_rate": 7.805800370178925e-06,
+      "loss": 0.2592,
+      "step": 3402
+    },
+    {
+      "epoch": 0.3311922141119221,
+      "grad_norm": 1.3572077520529662,
+      "learning_rate": 7.804495635586274e-06,
+      "loss": 0.3838,
+      "step": 3403
+    },
+    {
+      "epoch": 0.3312895377128954,
+      "grad_norm": 1.6103699950857218,
+      "learning_rate": 7.80319062230917e-06,
+      "loss": 0.2847,
+      "step": 3404
+    },
+    {
+      "epoch": 0.33138686131386863,
+      "grad_norm": 1.1427751941761943,
+      "learning_rate": 7.80188533047729e-06,
+      "loss": 0.3235,
+      "step": 3405
+    },
+    {
+      "epoch": 0.33148418491484183,
+      "grad_norm": 1.4205910051862616,
+      "learning_rate": 7.800579760220343e-06,
+      "loss": 0.4415,
+      "step": 3406
+    },
+    {
+      "epoch": 0.3315815085158151,
+      "grad_norm": 1.239833112068907,
+      "learning_rate": 7.799273911668062e-06,
+      "loss": 0.296,
+      "step": 3407
+    },
+    {
+      "epoch": 0.33167883211678834,
+      "grad_norm": 1.382006652854662,
+      "learning_rate": 7.797967784950215e-06,
+      "loss": 0.5129,
+      "step": 3408
+    },
+    {
+      "epoch": 0.33177615571776153,
+      "grad_norm": 1.3910482812533478,
+      "learning_rate": 7.796661380196587e-06,
+      "loss": 0.4355,
+      "step": 3409
+    },
+    {
+      "epoch": 0.3318734793187348,
+      "grad_norm": 1.3166393673557537,
+      "learning_rate": 7.795354697537e-06,
+      "loss": 0.3357,
+      "step": 3410
+    },
+    {
+      "epoch": 0.33197080291970804,
+      "grad_norm": 1.3540344564992455,
+      "learning_rate": 7.794047737101298e-06,
+      "loss": 0.2772,
+      "step": 3411
+    },
+    {
+      "epoch": 0.33206812652068124,
+      "grad_norm": 1.5732997739305445,
+      "learning_rate": 7.792740499019354e-06,
+      "loss": 0.368,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3321654501216545,
+      "grad_norm": 1.00398349093736,
+      "learning_rate": 7.791432983421071e-06,
+      "loss": 0.2794,
+      "step": 3413
+    },
+    {
+      "epoch": 0.33226277372262775,
+      "grad_norm": 1.5831231764140208,
+      "learning_rate": 7.790125190436378e-06,
+      "loss": 0.399,
+      "step": 3414
+    },
+    {
+      "epoch": 0.332360097323601,
+      "grad_norm": 1.301291609070449,
+      "learning_rate": 7.788817120195228e-06,
+      "loss": 0.4975,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3324574209245742,
+      "grad_norm": 3.04667201221868,
+      "learning_rate": 7.787508772827606e-06,
+      "loss": 0.3034,
+      "step": 3416
+    },
+    {
+      "epoch": 0.33255474452554745,
+      "grad_norm": 1.3711038355442808,
+      "learning_rate": 7.786200148463525e-06,
+      "loss": 0.4023,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3326520681265207,
+      "grad_norm": 1.4460757328108422,
+      "learning_rate": 7.784891247233025e-06,
+      "loss": 0.5218,
+      "step": 3418
+    },
+    {
+      "epoch": 0.3327493917274939,
+      "grad_norm": 1.5174415949182483,
+      "learning_rate": 7.783582069266167e-06,
+      "loss": 0.5401,
+      "step": 3419
+    },
+    {
+      "epoch": 0.33284671532846716,
+      "grad_norm": 1.170900270080405,
+      "learning_rate": 7.78227261469305e-06,
+      "loss": 0.3093,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3329440389294404,
+      "grad_norm": 1.1117784496155982,
+      "learning_rate": 7.78096288364379e-06,
+      "loss": 0.2159,
+      "step": 3421
+    },
+    {
+      "epoch": 0.3330413625304136,
+      "grad_norm": 1.385907002729564,
+      "learning_rate": 7.779652876248541e-06,
+      "loss": 0.4513,
+      "step": 3422
+    },
+    {
+      "epoch": 0.33313868613138686,
+      "grad_norm": 1.091845134289533,
+      "learning_rate": 7.778342592637477e-06,
+      "loss": 0.249,
+      "step": 3423
+    },
+    {
+      "epoch": 0.3332360097323601,
+      "grad_norm": 1.1987125682853903,
+      "learning_rate": 7.7770320329408e-06,
+      "loss": 0.4583,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.150260309114711,
+      "learning_rate": 7.775721197288746e-06,
+      "loss": 0.4145,
+      "step": 3425
+    },
+    {
+      "epoch": 0.33343065693430657,
+      "grad_norm": 1.1244746146994131,
+      "learning_rate": 7.77441008581157e-06,
+      "loss": 0.2334,
+      "step": 3426
+    },
+    {
+      "epoch": 0.3335279805352798,
+      "grad_norm": 2.372002969337908,
+      "learning_rate": 7.773098698639558e-06,
+      "loss": 0.3346,
+      "step": 3427
+    },
+    {
+      "epoch": 0.333625304136253,
+      "grad_norm": 1.460305633169593,
+      "learning_rate": 7.771787035903023e-06,
+      "loss": 0.5202,
+      "step": 3428
+    },
+    {
+      "epoch": 0.3337226277372263,
+      "grad_norm": 1.4552706392676258,
+      "learning_rate": 7.77047509773231e-06,
+      "loss": 0.3249,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3338199513381995,
+      "grad_norm": 26.423500051667432,
+      "learning_rate": 7.769162884257778e-06,
+      "loss": 0.3919,
+      "step": 3430
+    },
+    {
+      "epoch": 0.3339172749391727,
+      "grad_norm": 1.3364724766772538,
+      "learning_rate": 7.767850395609832e-06,
+      "loss": 0.4882,
+      "step": 3431
+    },
+    {
+      "epoch": 0.334014598540146,
+      "grad_norm": 1.162494089255124,
+      "learning_rate": 7.766537631918888e-06,
+      "loss": 0.4172,
+      "step": 3432
+    },
+    {
+      "epoch": 0.33411192214111923,
+      "grad_norm": 1.4584273535075323,
+      "learning_rate": 7.765224593315402e-06,
+      "loss": 0.5721,
+      "step": 3433
+    },
+    {
+      "epoch": 0.33420924574209243,
+      "grad_norm": 1.39936657097592,
+      "learning_rate": 7.763911279929848e-06,
+      "loss": 0.4454,
+      "step": 3434
+    },
+    {
+      "epoch": 0.3343065693430657,
+      "grad_norm": 1.5774498430067907,
+      "learning_rate": 7.76259769189273e-06,
+      "loss": 0.6756,
+      "step": 3435
+    },
+    {
+      "epoch": 0.33440389294403894,
+      "grad_norm": 1.4346477573335101,
+      "learning_rate": 7.761283829334583e-06,
+      "loss": 0.4939,
+      "step": 3436
+    },
+    {
+      "epoch": 0.3345012165450122,
+      "grad_norm": 1.2329772568374064,
+      "learning_rate": 7.759969692385963e-06,
+      "loss": 0.3576,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3345985401459854,
+      "grad_norm": 86.06815068595351,
+      "learning_rate": 7.75865528117746e-06,
+      "loss": 0.7983,
+      "step": 3438
+    },
+    {
+      "epoch": 0.33469586374695864,
+      "grad_norm": 1.4916748444459116,
+      "learning_rate": 7.757340595839686e-06,
+      "loss": 0.3408,
+      "step": 3439
+    },
+    {
+      "epoch": 0.3347931873479319,
+      "grad_norm": 1.204864631425379,
+      "learning_rate": 7.756025636503281e-06,
+      "loss": 0.2893,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3348905109489051,
+      "grad_norm": 1.1483309949418294,
+      "learning_rate": 7.754710403298915e-06,
+      "loss": 0.307,
+      "step": 3441
+    },
+    {
+      "epoch": 0.33498783454987835,
+      "grad_norm": 1.3801437746700074,
+      "learning_rate": 7.753394896357283e-06,
+      "loss": 0.5086,
+      "step": 3442
+    },
+    {
+      "epoch": 0.3350851581508516,
+      "grad_norm": 1.4670528589774587,
+      "learning_rate": 7.752079115809105e-06,
+      "loss": 0.5494,
+      "step": 3443
+    },
+    {
+      "epoch": 0.3351824817518248,
+      "grad_norm": 1.2268331435647832,
+      "learning_rate": 7.750763061785139e-06,
+      "loss": 0.3421,
+      "step": 3444
+    },
+    {
+      "epoch": 0.33527980535279805,
+      "grad_norm": 1.117498287907938,
+      "learning_rate": 7.749446734416153e-06,
+      "loss": 0.3583,
+      "step": 3445
+    },
+    {
+      "epoch": 0.3353771289537713,
+      "grad_norm": 1.6628933950216975,
+      "learning_rate": 7.748130133832956e-06,
+      "loss": 0.4265,
+      "step": 3446
+    },
+    {
+      "epoch": 0.3354744525547445,
+      "grad_norm": 1.4371941282513903,
+      "learning_rate": 7.746813260166379e-06,
+      "loss": 0.5499,
+      "step": 3447
+    },
+    {
+      "epoch": 0.33557177615571776,
+      "grad_norm": 1.4139636094726638,
+      "learning_rate": 7.74549611354728e-06,
+      "loss": 0.5113,
+      "step": 3448
+    },
+    {
+      "epoch": 0.335669099756691,
+      "grad_norm": 1.3607040572953095,
+      "learning_rate": 7.744178694106545e-06,
+      "loss": 0.3662,
+      "step": 3449
+    },
+    {
+      "epoch": 0.3357664233576642,
+      "grad_norm": 1.3497420382405303,
+      "learning_rate": 7.742861001975086e-06,
+      "loss": 0.37,
+      "step": 3450
+    },
+    {
+      "epoch": 0.33586374695863747,
+      "grad_norm": 1.1583812763163044,
+      "learning_rate": 7.741543037283844e-06,
+      "loss": 0.2328,
+      "step": 3451
+    },
+    {
+      "epoch": 0.3359610705596107,
+      "grad_norm": 1.233691835278808,
+      "learning_rate": 7.740224800163783e-06,
+      "loss": 0.3023,
+      "step": 3452
+    },
+    {
+      "epoch": 0.3360583941605839,
+      "grad_norm": 1.770041486932794,
+      "learning_rate": 7.738906290745902e-06,
+      "loss": 0.4559,
+      "step": 3453
+    },
+    {
+      "epoch": 0.33615571776155717,
+      "grad_norm": 1.6249038227669963,
+      "learning_rate": 7.737587509161218e-06,
+      "loss": 0.3305,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3362530413625304,
+      "grad_norm": 1.7123859840919058,
+      "learning_rate": 7.73626845554078e-06,
+      "loss": 0.8223,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3363503649635036,
+      "grad_norm": 1.465565821382168,
+      "learning_rate": 7.734949130015665e-06,
+      "loss": 0.3951,
+      "step": 3456
+    },
+    {
+      "epoch": 0.3364476885644769,
+      "grad_norm": 1.3358312642650005,
+      "learning_rate": 7.733629532716974e-06,
+      "loss": 0.3988,
+      "step": 3457
+    },
+    {
+      "epoch": 0.33654501216545013,
+      "grad_norm": 1.4614411980006665,
+      "learning_rate": 7.732309663775834e-06,
+      "loss": 0.4447,
+      "step": 3458
+    },
+    {
+      "epoch": 0.3366423357664234,
+      "grad_norm": 1.5007446348141111,
+      "learning_rate": 7.730989523323405e-06,
+      "loss": 0.5075,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3367396593673966,
+      "grad_norm": 1.378374467272079,
+      "learning_rate": 7.72966911149087e-06,
+      "loss": 0.3713,
+      "step": 3460
+    },
+    {
+      "epoch": 0.33683698296836984,
+      "grad_norm": 1.264554548276351,
+      "learning_rate": 7.728348428409434e-06,
+      "loss": 0.4239,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3369343065693431,
+      "grad_norm": 1.3654939025524866,
+      "learning_rate": 7.72702747421034e-06,
+      "loss": 0.3861,
+      "step": 3462
+    },
+    {
+      "epoch": 0.3370316301703163,
+      "grad_norm": 1.456076628879786,
+      "learning_rate": 7.72570624902485e-06,
+      "loss": 0.4548,
+      "step": 3463
+    },
+    {
+      "epoch": 0.33712895377128954,
+      "grad_norm": 1.5979907891076075,
+      "learning_rate": 7.724384752984253e-06,
+      "loss": 0.5395,
+      "step": 3464
+    },
+    {
+      "epoch": 0.3372262773722628,
+      "grad_norm": 1.479690145713753,
+      "learning_rate": 7.723062986219871e-06,
+      "loss": 0.4676,
+      "step": 3465
+    },
+    {
+      "epoch": 0.337323600973236,
+      "grad_norm": 1.5794398807050158,
+      "learning_rate": 7.721740948863044e-06,
+      "loss": 0.6383,
+      "step": 3466
+    },
+    {
+      "epoch": 0.33742092457420925,
+      "grad_norm": 1.5077566027780562,
+      "learning_rate": 7.720418641045147e-06,
+      "loss": 0.449,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3375182481751825,
+      "grad_norm": 1.536687422230877,
+      "learning_rate": 7.719096062897578e-06,
+      "loss": 0.3885,
+      "step": 3468
+    },
+    {
+      "epoch": 0.3376155717761557,
+      "grad_norm": 1.4136867972495795,
+      "learning_rate": 7.717773214551762e-06,
+      "loss": 0.4262,
+      "step": 3469
+    },
+    {
+      "epoch": 0.33771289537712895,
+      "grad_norm": 1.7521028146405362,
+      "learning_rate": 7.71645009613915e-06,
+      "loss": 0.3553,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3378102189781022,
+      "grad_norm": 1.28472614917785,
+      "learning_rate": 7.715126707791223e-06,
+      "loss": 0.5044,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3379075425790754,
+      "grad_norm": 1.5285157685020219,
+      "learning_rate": 7.713803049639485e-06,
+      "loss": 0.3067,
+      "step": 3472
+    },
+    {
+      "epoch": 0.33800486618004866,
+      "grad_norm": 1.301311213128161,
+      "learning_rate": 7.712479121815473e-06,
+      "loss": 0.4036,
+      "step": 3473
+    },
+    {
+      "epoch": 0.3381021897810219,
+      "grad_norm": 1.4164505494518185,
+      "learning_rate": 7.711154924450741e-06,
+      "loss": 0.3674,
+      "step": 3474
+    },
+    {
+      "epoch": 0.3381995133819951,
+      "grad_norm": 1.2252498333021546,
+      "learning_rate": 7.709830457676876e-06,
+      "loss": 0.273,
+      "step": 3475
+    },
+    {
+      "epoch": 0.33829683698296836,
+      "grad_norm": 1.4131843411362779,
+      "learning_rate": 7.708505721625497e-06,
+      "loss": 0.565,
+      "step": 3476
+    },
+    {
+      "epoch": 0.3383941605839416,
+      "grad_norm": 1.6947767404353455,
+      "learning_rate": 7.707180716428237e-06,
+      "loss": 0.4248,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3384914841849148,
+      "grad_norm": 1.2913551057356365,
+      "learning_rate": 7.705855442216766e-06,
+      "loss": 0.3537,
+      "step": 3478
+    },
+    {
+      "epoch": 0.33858880778588807,
+      "grad_norm": 1.150600446597589,
+      "learning_rate": 7.704529899122776e-06,
+      "loss": 0.3311,
+      "step": 3479
+    },
+    {
+      "epoch": 0.3386861313868613,
+      "grad_norm": 1.2021311829209522,
+      "learning_rate": 7.703204087277989e-06,
+      "loss": 0.4163,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3387834549878346,
+      "grad_norm": 1.256321657329123,
+      "learning_rate": 7.70187800681415e-06,
+      "loss": 0.3609,
+      "step": 3481
+    },
+    {
+      "epoch": 0.3388807785888078,
+      "grad_norm": 1.3378074898611492,
+      "learning_rate": 7.70055165786303e-06,
+      "loss": 0.3365,
+      "step": 3482
+    },
+    {
+      "epoch": 0.338978102189781,
+      "grad_norm": 1.170985634605712,
+      "learning_rate": 7.699225040556435e-06,
+      "loss": 0.2524,
+      "step": 3483
+    },
+    {
+      "epoch": 0.3390754257907543,
+      "grad_norm": 1.3653491057947706,
+      "learning_rate": 7.697898155026188e-06,
+      "loss": 0.282,
+      "step": 3484
+    },
+    {
+      "epoch": 0.3391727493917275,
+      "grad_norm": 1.3228454670807173,
+      "learning_rate": 7.696571001404143e-06,
+      "loss": 0.4489,
+      "step": 3485
+    },
+    {
+      "epoch": 0.33927007299270073,
+      "grad_norm": 1.568456967406488,
+      "learning_rate": 7.695243579822179e-06,
+      "loss": 0.672,
+      "step": 3486
+    },
+    {
+      "epoch": 0.339367396593674,
+      "grad_norm": 1.5380107558510523,
+      "learning_rate": 7.693915890412205e-06,
+      "loss": 0.4099,
+      "step": 3487
+    },
+    {
+      "epoch": 0.3394647201946472,
+      "grad_norm": 1.522434941547624,
+      "learning_rate": 7.692587933306152e-06,
+      "loss": 0.3895,
+      "step": 3488
+    },
+    {
+      "epoch": 0.33956204379562044,
+      "grad_norm": 1.6424350744273293,
+      "learning_rate": 7.691259708635983e-06,
+      "loss": 0.4547,
+      "step": 3489
+    },
+    {
+      "epoch": 0.3396593673965937,
+      "grad_norm": 1.1285195925950828,
+      "learning_rate": 7.689931216533682e-06,
+      "loss": 0.3205,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3397566909975669,
+      "grad_norm": 1.2211335178765037,
+      "learning_rate": 7.68860245713126e-06,
+      "loss": 0.2849,
+      "step": 3491
+    },
+    {
+      "epoch": 0.33985401459854014,
+      "grad_norm": 1.3440619705365895,
+      "learning_rate": 7.687273430560763e-06,
+      "loss": 0.3493,
+      "step": 3492
+    },
+    {
+      "epoch": 0.3399513381995134,
+      "grad_norm": 1.9270777796759784,
+      "learning_rate": 7.685944136954252e-06,
+      "loss": 0.3207,
+      "step": 3493
+    },
+    {
+      "epoch": 0.3400486618004866,
+      "grad_norm": 1.3294429746967642,
+      "learning_rate": 7.684614576443821e-06,
+      "loss": 0.3215,
+      "step": 3494
+    },
+    {
+      "epoch": 0.34014598540145985,
+      "grad_norm": 1.299183824061712,
+      "learning_rate": 7.68328474916159e-06,
+      "loss": 0.3565,
+      "step": 3495
+    },
+    {
+      "epoch": 0.3402433090024331,
+      "grad_norm": 1.5283573621457682,
+      "learning_rate": 7.681954655239703e-06,
+      "loss": 0.4789,
+      "step": 3496
+    },
+    {
+      "epoch": 0.3403406326034063,
+      "grad_norm": 1.4764556038728422,
+      "learning_rate": 7.680624294810335e-06,
+      "loss": 0.4079,
+      "step": 3497
+    },
+    {
+      "epoch": 0.34043795620437955,
+      "grad_norm": 1.5139159959394148,
+      "learning_rate": 7.679293668005683e-06,
+      "loss": 0.5341,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3405352798053528,
+      "grad_norm": 1.5537302722123498,
+      "learning_rate": 7.677962774957971e-06,
+      "loss": 0.478,
+      "step": 3499
+    },
+    {
+      "epoch": 0.340632603406326,
+      "grad_norm": 1.5827415849591213,
+      "learning_rate": 7.676631615799453e-06,
+      "loss": 0.359,
+      "step": 3500
+    },
+    {
+      "epoch": 0.34072992700729926,
+      "grad_norm": 1.8389434879657838,
+      "learning_rate": 7.675300190662406e-06,
+      "loss": 0.3688,
+      "step": 3501
+    },
+    {
+      "epoch": 0.3408272506082725,
+      "grad_norm": 1.219949474382852,
+      "learning_rate": 7.673968499679134e-06,
+      "loss": 0.3099,
+      "step": 3502
+    },
+    {
+      "epoch": 0.34092457420924577,
+      "grad_norm": 1.5256977236182008,
+      "learning_rate": 7.67263654298197e-06,
+      "loss": 0.3838,
+      "step": 3503
+    },
+    {
+      "epoch": 0.34102189781021897,
+      "grad_norm": 1.4659409139833723,
+      "learning_rate": 7.671304320703269e-06,
+      "loss": 0.4845,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3411192214111922,
+      "grad_norm": 1.667843387368496,
+      "learning_rate": 7.669971832975417e-06,
+      "loss": 0.5876,
+      "step": 3505
+    },
+    {
+      "epoch": 0.3412165450121655,
+      "grad_norm": 1.346414364244355,
+      "learning_rate": 7.668639079930821e-06,
+      "loss": 0.4337,
+      "step": 3506
+    },
+    {
+      "epoch": 0.34131386861313867,
+      "grad_norm": 2.022273962642171,
+      "learning_rate": 7.66730606170192e-06,
+      "loss": 0.472,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3414111922141119,
+      "grad_norm": 1.4021904736753643,
+      "learning_rate": 7.665972778421175e-06,
+      "loss": 0.3331,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3415085158150852,
+      "grad_norm": 1.3964469851310124,
+      "learning_rate": 7.664639230221081e-06,
+      "loss": 0.4151,
+      "step": 3509
+    },
+    {
+      "epoch": 0.3416058394160584,
+      "grad_norm": 1.437167087649688,
+      "learning_rate": 7.663305417234146e-06,
+      "loss": 0.3751,
+      "step": 3510
+    },
+    {
+      "epoch": 0.34170316301703163,
+      "grad_norm": 1.3813486918408102,
+      "learning_rate": 7.661971339592913e-06,
+      "loss": 0.2492,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3418004866180049,
+      "grad_norm": 2.213050506553265,
+      "learning_rate": 7.660636997429953e-06,
+      "loss": 0.2442,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3418978102189781,
+      "grad_norm": 1.4125007730667565,
+      "learning_rate": 7.659302390877858e-06,
+      "loss": 0.3901,
+      "step": 3513
+    },
+    {
+      "epoch": 0.34199513381995134,
+      "grad_norm": 1.5262547193449572,
+      "learning_rate": 7.657967520069253e-06,
+      "loss": 0.5142,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3420924574209246,
+      "grad_norm": 1.419026298338398,
+      "learning_rate": 7.65663238513678e-06,
+      "loss": 0.4604,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3421897810218978,
+      "grad_norm": 1.354197849012477,
+      "learning_rate": 7.655296986213114e-06,
+      "loss": 0.3741,
+      "step": 3516
+    },
+    {
+      "epoch": 0.34228710462287104,
+      "grad_norm": 1.3675444218814583,
+      "learning_rate": 7.653961323430954e-06,
+      "loss": 0.4636,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3423844282238443,
+      "grad_norm": 1.6737094690203054,
+      "learning_rate": 7.652625396923027e-06,
+      "loss": 0.5368,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3424817518248175,
+      "grad_norm": 1.5346015529843957,
+      "learning_rate": 7.651289206822084e-06,
+      "loss": 0.5808,
+      "step": 3519
+    },
+    {
+      "epoch": 0.34257907542579075,
+      "grad_norm": 1.7677803050375525,
+      "learning_rate": 7.649952753260901e-06,
+      "loss": 0.4331,
+      "step": 3520
+    },
+    {
+      "epoch": 0.342676399026764,
+      "grad_norm": 1.5130877149006923,
+      "learning_rate": 7.648616036372288e-06,
+      "loss": 0.4343,
+      "step": 3521
+    },
+    {
+      "epoch": 0.34277372262773725,
+      "grad_norm": 1.6505485143132894,
+      "learning_rate": 7.647279056289068e-06,
+      "loss": 0.4256,
+      "step": 3522
+    },
+    {
+      "epoch": 0.34287104622871045,
+      "grad_norm": 1.1213144756308453,
+      "learning_rate": 7.6459418131441e-06,
+      "loss": 0.248,
+      "step": 3523
+    },
+    {
+      "epoch": 0.3429683698296837,
+      "grad_norm": 1.4978011226878554,
+      "learning_rate": 7.64460430707027e-06,
+      "loss": 0.4457,
+      "step": 3524
+    },
+    {
+      "epoch": 0.34306569343065696,
+      "grad_norm": 1.724927585574558,
+      "learning_rate": 7.643266538200485e-06,
+      "loss": 0.5746,
+      "step": 3525
+    },
+    {
+      "epoch": 0.34316301703163016,
+      "grad_norm": 1.4218186587395187,
+      "learning_rate": 7.641928506667677e-06,
+      "loss": 0.4628,
+      "step": 3526
+    },
+    {
+      "epoch": 0.3432603406326034,
+      "grad_norm": 1.1979163238967183,
+      "learning_rate": 7.640590212604813e-06,
+      "loss": 0.3276,
+      "step": 3527
+    },
+    {
+      "epoch": 0.34335766423357666,
+      "grad_norm": 1.5459257353579257,
+      "learning_rate": 7.639251656144873e-06,
+      "loss": 0.5543,
+      "step": 3528
+    },
+    {
+      "epoch": 0.34345498783454986,
+      "grad_norm": 2.168708523840157,
+      "learning_rate": 7.637912837420876e-06,
+      "loss": 0.4451,
+      "step": 3529
+    },
+    {
+      "epoch": 0.3435523114355231,
+      "grad_norm": 1.3295572972455665,
+      "learning_rate": 7.63657375656586e-06,
+      "loss": 0.3659,
+      "step": 3530
+    },
+    {
+      "epoch": 0.34364963503649637,
+      "grad_norm": 1.3940981237720802,
+      "learning_rate": 7.635234413712886e-06,
+      "loss": 0.3305,
+      "step": 3531
+    },
+    {
+      "epoch": 0.34374695863746957,
+      "grad_norm": 1.668696796269181,
+      "learning_rate": 7.63389480899505e-06,
+      "loss": 0.262,
+      "step": 3532
+    },
+    {
+      "epoch": 0.3438442822384428,
+      "grad_norm": 1.3235762846506705,
+      "learning_rate": 7.632554942545468e-06,
+      "loss": 0.354,
+      "step": 3533
+    },
+    {
+      "epoch": 0.3439416058394161,
+      "grad_norm": 1.774595647074867,
+      "learning_rate": 7.631214814497283e-06,
+      "loss": 0.4181,
+      "step": 3534
+    },
+    {
+      "epoch": 0.3440389294403893,
+      "grad_norm": 1.3699960689070612,
+      "learning_rate": 7.629874424983664e-06,
+      "loss": 0.4893,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3441362530413625,
+      "grad_norm": 1.0054517790481798,
+      "learning_rate": 7.628533774137809e-06,
+      "loss": 0.2678,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3442335766423358,
+      "grad_norm": 1.2585804176689974,
+      "learning_rate": 7.627192862092936e-06,
+      "loss": 0.3145,
+      "step": 3537
+    },
+    {
+      "epoch": 0.344330900243309,
+      "grad_norm": 1.4486458769434574,
+      "learning_rate": 7.625851688982293e-06,
+      "loss": 0.5018,
+      "step": 3538
+    },
+    {
+      "epoch": 0.34442822384428223,
+      "grad_norm": 1.6761010493950546,
+      "learning_rate": 7.624510254939155e-06,
+      "loss": 0.5786,
+      "step": 3539
+    },
+    {
+      "epoch": 0.3445255474452555,
+      "grad_norm": 1.1326870552848924,
+      "learning_rate": 7.623168560096819e-06,
+      "loss": 0.2714,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3446228710462287,
+      "grad_norm": 1.380514271735255,
+      "learning_rate": 7.62182660458861e-06,
+      "loss": 0.4435,
+      "step": 3541
+    },
+    {
+      "epoch": 0.34472019464720194,
+      "grad_norm": 1.3847938610044406,
+      "learning_rate": 7.620484388547881e-06,
+      "loss": 0.456,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3448175182481752,
+      "grad_norm": 1.4774572497281164,
+      "learning_rate": 7.619141912108008e-06,
+      "loss": 0.4016,
+      "step": 3543
+    },
+    {
+      "epoch": 0.34491484184914845,
+      "grad_norm": 1.4155087320104913,
+      "learning_rate": 7.617799175402392e-06,
+      "loss": 0.4672,
+      "step": 3544
+    },
+    {
+      "epoch": 0.34501216545012164,
+      "grad_norm": 1.2347049030843364,
+      "learning_rate": 7.616456178564463e-06,
+      "loss": 0.4701,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3451094890510949,
+      "grad_norm": 1.3617965352115597,
+      "learning_rate": 7.615112921727677e-06,
+      "loss": 0.4411,
+      "step": 3546
+    },
+    {
+      "epoch": 0.34520681265206815,
+      "grad_norm": 1.5080403368283026,
+      "learning_rate": 7.613769405025511e-06,
+      "loss": 0.446,
+      "step": 3547
+    },
+    {
+      "epoch": 0.34530413625304135,
+      "grad_norm": 1.3587460871748047,
+      "learning_rate": 7.612425628591473e-06,
+      "loss": 0.4618,
+      "step": 3548
+    },
+    {
+      "epoch": 0.3454014598540146,
+      "grad_norm": 1.5594894397765757,
+      "learning_rate": 7.611081592559095e-06,
+      "loss": 0.6454,
+      "step": 3549
+    },
+    {
+      "epoch": 0.34549878345498786,
+      "grad_norm": 1.5472776871393785,
+      "learning_rate": 7.609737297061934e-06,
+      "loss": 0.4209,
+      "step": 3550
+    },
+    {
+      "epoch": 0.34559610705596105,
+      "grad_norm": 1.1387925779252865,
+      "learning_rate": 7.608392742233573e-06,
+      "loss": 0.2542,
+      "step": 3551
+    },
+    {
+      "epoch": 0.3456934306569343,
+      "grad_norm": 2.1308610533262753,
+      "learning_rate": 7.6070479282076226e-06,
+      "loss": 0.4232,
+      "step": 3552
+    },
+    {
+      "epoch": 0.34579075425790756,
+      "grad_norm": 1.655772230226882,
+      "learning_rate": 7.605702855117717e-06,
+      "loss": 0.316,
+      "step": 3553
+    },
+    {
+      "epoch": 0.34588807785888076,
+      "grad_norm": 1.2773133253854652,
+      "learning_rate": 7.604357523097518e-06,
+      "loss": 0.3933,
+      "step": 3554
+    },
+    {
+      "epoch": 0.345985401459854,
+      "grad_norm": 1.5157476945746606,
+      "learning_rate": 7.6030119322807105e-06,
+      "loss": 0.4895,
+      "step": 3555
+    },
+    {
+      "epoch": 0.34608272506082727,
+      "grad_norm": 1.5384836415390328,
+      "learning_rate": 7.601666082801007e-06,
+      "loss": 0.4571,
+      "step": 3556
+    },
+    {
+      "epoch": 0.34618004866180047,
+      "grad_norm": 1.3494696317056387,
+      "learning_rate": 7.600319974792145e-06,
+      "loss": 0.3585,
+      "step": 3557
+    },
+    {
+      "epoch": 0.3462773722627737,
+      "grad_norm": 1.5639692646479821,
+      "learning_rate": 7.59897360838789e-06,
+      "loss": 0.4913,
+      "step": 3558
+    },
+    {
+      "epoch": 0.346374695863747,
+      "grad_norm": 1.5771201683321747,
+      "learning_rate": 7.59762698372203e-06,
+      "loss": 0.644,
+      "step": 3559
+    },
+    {
+      "epoch": 0.34647201946472017,
+      "grad_norm": 1.445755844165086,
+      "learning_rate": 7.596280100928379e-06,
+      "loss": 0.4662,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3465693430656934,
+      "grad_norm": 1.0169653044814775,
+      "learning_rate": 7.59493296014078e-06,
+      "loss": 0.1873,
+      "step": 3561
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 1.3573861316393436,
+      "learning_rate": 7.593585561493098e-06,
+      "loss": 0.3621,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3467639902676399,
+      "grad_norm": 1.3679634725925294,
+      "learning_rate": 7.592237905119224e-06,
+      "loss": 0.3714,
+      "step": 3563
+    },
+    {
+      "epoch": 0.34686131386861313,
+      "grad_norm": 1.6153262572349345,
+      "learning_rate": 7.590889991153076e-06,
+      "loss": 0.1934,
+      "step": 3564
+    },
+    {
+      "epoch": 0.3469586374695864,
+      "grad_norm": 1.3501706757209104,
+      "learning_rate": 7.589541819728597e-06,
+      "loss": 0.3771,
+      "step": 3565
+    },
+    {
+      "epoch": 0.34705596107055964,
+      "grad_norm": 1.4359834383485004,
+      "learning_rate": 7.588193390979756e-06,
+      "loss": 0.4021,
+      "step": 3566
+    },
+    {
+      "epoch": 0.34715328467153284,
+      "grad_norm": 1.6178918309073458,
+      "learning_rate": 7.5868447050405456e-06,
+      "loss": 0.5326,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3472506082725061,
+      "grad_norm": 1.2161304982080974,
+      "learning_rate": 7.585495762044989e-06,
+      "loss": 0.3215,
+      "step": 3568
+    },
+    {
+      "epoch": 0.34734793187347934,
+      "grad_norm": 1.2514780842366786,
+      "learning_rate": 7.584146562127128e-06,
+      "loss": 0.2619,
+      "step": 3569
+    },
+    {
+      "epoch": 0.34744525547445254,
+      "grad_norm": 1.292347877079837,
+      "learning_rate": 7.5827971054210334e-06,
+      "loss": 0.3722,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3475425790754258,
+      "grad_norm": 1.428500991174703,
+      "learning_rate": 7.581447392060806e-06,
+      "loss": 0.4681,
+      "step": 3571
+    },
+    {
+      "epoch": 0.34763990267639905,
+      "grad_norm": 1.6187169310973553,
+      "learning_rate": 7.5800974221805635e-06,
+      "loss": 0.3123,
+      "step": 3572
+    },
+    {
+      "epoch": 0.34773722627737225,
+      "grad_norm": 1.2828962403261588,
+      "learning_rate": 7.5787471959144535e-06,
+      "loss": 0.3426,
+      "step": 3573
+    },
+    {
+      "epoch": 0.3478345498783455,
+      "grad_norm": 1.3424794600068044,
+      "learning_rate": 7.577396713396649e-06,
+      "loss": 0.2749,
+      "step": 3574
+    },
+    {
+      "epoch": 0.34793187347931875,
+      "grad_norm": 1.311429995391481,
+      "learning_rate": 7.576045974761352e-06,
+      "loss": 0.1912,
+      "step": 3575
+    },
+    {
+      "epoch": 0.34802919708029195,
+      "grad_norm": 1.9475934853598964,
+      "learning_rate": 7.57469498014278e-06,
+      "loss": 0.3568,
+      "step": 3576
+    },
+    {
+      "epoch": 0.3481265206812652,
+      "grad_norm": 1.5712590946732445,
+      "learning_rate": 7.573343729675187e-06,
+      "loss": 0.4361,
+      "step": 3577
+    },
+    {
+      "epoch": 0.34822384428223846,
+      "grad_norm": 1.3460143771601136,
+      "learning_rate": 7.5719922234928435e-06,
+      "loss": 0.3258,
+      "step": 3578
+    },
+    {
+      "epoch": 0.34832116788321166,
+      "grad_norm": 1.3457032135851295,
+      "learning_rate": 7.5706404617300544e-06,
+      "loss": 0.2679,
+      "step": 3579
+    },
+    {
+      "epoch": 0.3484184914841849,
+      "grad_norm": 1.602637400100207,
+      "learning_rate": 7.569288444521141e-06,
+      "loss": 0.4349,
+      "step": 3580
+    },
+    {
+      "epoch": 0.34851581508515816,
+      "grad_norm": 1.193898879748489,
+      "learning_rate": 7.567936172000456e-06,
+      "loss": 0.3823,
+      "step": 3581
+    },
+    {
+      "epoch": 0.34861313868613136,
+      "grad_norm": 1.7213934943419156,
+      "learning_rate": 7.5665836443023764e-06,
+      "loss": 0.4477,
+      "step": 3582
+    },
+    {
+      "epoch": 0.3487104622871046,
+      "grad_norm": 2.0896889317289262,
+      "learning_rate": 7.5652308615613025e-06,
+      "loss": 0.548,
+      "step": 3583
+    },
+    {
+      "epoch": 0.34880778588807787,
+      "grad_norm": 1.2133584709672542,
+      "learning_rate": 7.563877823911661e-06,
+      "loss": 0.3342,
+      "step": 3584
+    },
+    {
+      "epoch": 0.34890510948905107,
+      "grad_norm": 1.3518803834565474,
+      "learning_rate": 7.562524531487902e-06,
+      "loss": 0.4428,
+      "step": 3585
+    },
+    {
+      "epoch": 0.3490024330900243,
+      "grad_norm": 1.5646028482749865,
+      "learning_rate": 7.561170984424509e-06,
+      "loss": 0.4805,
+      "step": 3586
+    },
+    {
+      "epoch": 0.3490997566909976,
+      "grad_norm": 1.725109749219042,
+      "learning_rate": 7.5598171828559775e-06,
+      "loss": 0.3953,
+      "step": 3587
+    },
+    {
+      "epoch": 0.34919708029197083,
+      "grad_norm": 1.6751385503374157,
+      "learning_rate": 7.558463126916842e-06,
+      "loss": 0.3466,
+      "step": 3588
+    },
+    {
+      "epoch": 0.349294403892944,
+      "grad_norm": 1.577501457080682,
+      "learning_rate": 7.557108816741651e-06,
+      "loss": 0.387,
+      "step": 3589
+    },
+    {
+      "epoch": 0.3493917274939173,
+      "grad_norm": 1.5319961753926956,
+      "learning_rate": 7.5557542524649866e-06,
+      "loss": 0.2916,
+      "step": 3590
+    },
+    {
+      "epoch": 0.34948905109489053,
+      "grad_norm": 2.5209551905243845,
+      "learning_rate": 7.554399434221449e-06,
+      "loss": 0.2941,
+      "step": 3591
+    },
+    {
+      "epoch": 0.34958637469586373,
+      "grad_norm": 1.1980216546975295,
+      "learning_rate": 7.553044362145672e-06,
+      "loss": 0.3867,
+      "step": 3592
+    },
+    {
+      "epoch": 0.349683698296837,
+      "grad_norm": 1.425370917650529,
+      "learning_rate": 7.551689036372306e-06,
+      "loss": 0.4788,
+      "step": 3593
+    },
+    {
+      "epoch": 0.34978102189781024,
+      "grad_norm": 1.5504040775883203,
+      "learning_rate": 7.550333457036032e-06,
+      "loss": 0.5355,
+      "step": 3594
+    },
+    {
+      "epoch": 0.34987834549878344,
+      "grad_norm": 1.253372887044746,
+      "learning_rate": 7.5489776242715564e-06,
+      "loss": 0.4783,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3499756690997567,
+      "grad_norm": 1.4985868632126829,
+      "learning_rate": 7.547621538213607e-06,
+      "loss": 0.4225,
+      "step": 3596
+    },
+    {
+      "epoch": 0.35007299270072995,
+      "grad_norm": 1.4435970539146459,
+      "learning_rate": 7.5462651989969385e-06,
+      "loss": 0.3745,
+      "step": 3597
+    },
+    {
+      "epoch": 0.35017031630170314,
+      "grad_norm": 1.3981266208958172,
+      "learning_rate": 7.5449086067563314e-06,
+      "loss": 0.3456,
+      "step": 3598
+    },
+    {
+      "epoch": 0.3502676399026764,
+      "grad_norm": 1.8315759023798575,
+      "learning_rate": 7.543551761626594e-06,
+      "loss": 0.5542,
+      "step": 3599
+    },
+    {
+      "epoch": 0.35036496350364965,
+      "grad_norm": 1.0610077420382762,
+      "learning_rate": 7.542194663742553e-06,
+      "loss": 0.3009,
+      "step": 3600
+    },
+    {
+      "epoch": 0.35046228710462285,
+      "grad_norm": 1.4599599576710758,
+      "learning_rate": 7.5408373132390674e-06,
+      "loss": 0.3322,
+      "step": 3601
+    },
+    {
+      "epoch": 0.3505596107055961,
+      "grad_norm": 1.357290752856556,
+      "learning_rate": 7.539479710251014e-06,
+      "loss": 0.4219,
+      "step": 3602
+    },
+    {
+      "epoch": 0.35065693430656936,
+      "grad_norm": 1.3466674715612543,
+      "learning_rate": 7.538121854913303e-06,
+      "loss": 0.4688,
+      "step": 3603
+    },
+    {
+      "epoch": 0.35075425790754255,
+      "grad_norm": 1.9207221814789595,
+      "learning_rate": 7.536763747360863e-06,
+      "loss": 0.5269,
+      "step": 3604
+    },
+    {
+      "epoch": 0.3508515815085158,
+      "grad_norm": 1.4616085693980927,
+      "learning_rate": 7.535405387728649e-06,
+      "loss": 0.5216,
+      "step": 3605
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 10275,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 515,
+  "total_flos": 461160819605504.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}