ViCA-20p / trainer_state.json
nkkbr's picture
Initial commit
7feb823
Raw
History Blame Contribute Delete
359 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.20048661800486617,
"eval_steps": 500,
"global_step": 2060,
"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
}
],
"logging_steps": 1.0,
"max_steps": 10275,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 515,
"total_flos": 263504598163456.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}