Files
DReP-SFT-Qwen3-4B-Thinking-…/trainer_state.json

5812 lines
140 KiB
JSON
Raw Permalink Normal View History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 500,
"global_step": 824,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00486470051687443,
"grad_norm": 10.097384452819824,
"learning_rate": 0.0,
"loss": 1.0734,
"step": 1
},
{
"epoch": 0.00972940103374886,
"grad_norm": 9.055700302124023,
"learning_rate": 1.204819277108434e-07,
"loss": 1.0536,
"step": 2
},
{
"epoch": 0.01459410155062329,
"grad_norm": 9.132233619689941,
"learning_rate": 2.409638554216868e-07,
"loss": 1.0409,
"step": 3
},
{
"epoch": 0.01945880206749772,
"grad_norm": 9.571680068969727,
"learning_rate": 3.614457831325301e-07,
"loss": 1.0465,
"step": 4
},
{
"epoch": 0.024323502584372148,
"grad_norm": 9.333101272583008,
"learning_rate": 4.819277108433736e-07,
"loss": 1.0507,
"step": 5
},
{
"epoch": 0.02918820310124658,
"grad_norm": 9.378636360168457,
"learning_rate": 6.024096385542169e-07,
"loss": 1.0471,
"step": 6
},
{
"epoch": 0.03405290361812101,
"grad_norm": 9.336066246032715,
"learning_rate": 7.228915662650602e-07,
"loss": 1.0409,
"step": 7
},
{
"epoch": 0.03891760413499544,
"grad_norm": 8.894888877868652,
"learning_rate": 8.433734939759036e-07,
"loss": 1.048,
"step": 8
},
{
"epoch": 0.04378230465186987,
"grad_norm": 8.849226951599121,
"learning_rate": 9.638554216867472e-07,
"loss": 1.0159,
"step": 9
},
{
"epoch": 0.048647005168744296,
"grad_norm": 8.914710998535156,
"learning_rate": 1.0843373493975905e-06,
"loss": 1.0504,
"step": 10
},
{
"epoch": 0.05351170568561873,
"grad_norm": 7.982012748718262,
"learning_rate": 1.2048192771084338e-06,
"loss": 0.9808,
"step": 11
},
{
"epoch": 0.05837640620249316,
"grad_norm": 7.971729755401611,
"learning_rate": 1.3253012048192773e-06,
"loss": 0.9917,
"step": 12
},
{
"epoch": 0.06324110671936758,
"grad_norm": 5.418062210083008,
"learning_rate": 1.4457831325301204e-06,
"loss": 0.9384,
"step": 13
},
{
"epoch": 0.06810580723624202,
"grad_norm": 5.036281108856201,
"learning_rate": 1.566265060240964e-06,
"loss": 0.9349,
"step": 14
},
{
"epoch": 0.07297050775311645,
"grad_norm": 4.816902160644531,
"learning_rate": 1.6867469879518073e-06,
"loss": 0.9049,
"step": 15
},
{
"epoch": 0.07783520826999088,
"grad_norm": 4.0932536125183105,
"learning_rate": 1.8072289156626508e-06,
"loss": 0.9008,
"step": 16
},
{
"epoch": 0.08269990878686531,
"grad_norm": 3.8836312294006348,
"learning_rate": 1.9277108433734943e-06,
"loss": 0.8729,
"step": 17
},
{
"epoch": 0.08756460930373974,
"grad_norm": 2.254310131072998,
"learning_rate": 2.0481927710843377e-06,
"loss": 0.8257,
"step": 18
},
{
"epoch": 0.09242930982061417,
"grad_norm": 2.2753942012786865,
"learning_rate": 2.168674698795181e-06,
"loss": 0.8689,
"step": 19
},
{
"epoch": 0.09729401033748859,
"grad_norm": 2.035552978515625,
"learning_rate": 2.2891566265060243e-06,
"loss": 0.8125,
"step": 20
},
{
"epoch": 0.10215871085436302,
"grad_norm": 1.9866139888763428,
"learning_rate": 2.4096385542168676e-06,
"loss": 0.843,
"step": 21
},
{
"epoch": 0.10702341137123746,
"grad_norm": 1.855616807937622,
"learning_rate": 2.530120481927711e-06,
"loss": 0.8453,
"step": 22
},
{
"epoch": 0.11188811188811189,
"grad_norm": 1.6492997407913208,
"learning_rate": 2.6506024096385547e-06,
"loss": 0.8286,
"step": 23
},
{
"epoch": 0.11675281240498632,
"grad_norm": 1.6260086297988892,
"learning_rate": 2.771084337349398e-06,
"loss": 0.815,
"step": 24
},
{
"epoch": 0.12161751292186075,
"grad_norm": 1.7872730493545532,
"learning_rate": 2.891566265060241e-06,
"loss": 0.8106,
"step": 25
},
{
"epoch": 0.12648221343873517,
"grad_norm": 1.858306884765625,
"learning_rate": 3.012048192771085e-06,
"loss": 0.7879,
"step": 26
},
{
"epoch": 0.1313469139556096,
"grad_norm": 1.847562551498413,
"learning_rate": 3.132530120481928e-06,
"loss": 0.79,
"step": 27
},
{
"epoch": 0.13621161447248403,
"grad_norm": 1.624657154083252,
"learning_rate": 3.2530120481927713e-06,
"loss": 0.7572,
"step": 28
},
{
"epoch": 0.14107631498935846,
"grad_norm": 1.4090893268585205,
"learning_rate": 3.3734939759036146e-06,
"loss": 0.7544,
"step": 29
},
{
"epoch": 0.1459410155062329,
"grad_norm": 1.184046745300293,
"learning_rate": 3.4939759036144583e-06,
"loss": 0.7419,
"step": 30
},
{
"epoch": 0.15080571602310733,
"grad_norm": 0.9826652407646179,
"learning_rate": 3.6144578313253016e-06,
"loss": 0.7562,
"step": 31
},
{
"epoch": 0.15567041653998176,
"grad_norm": 0.730877161026001,
"learning_rate": 3.7349397590361445e-06,
"loss": 0.7563,
"step": 32
},
{
"epoch": 0.1605351170568562,
"grad_norm": 0.672069251537323,
"learning_rate": 3.855421686746989e-06,
"loss": 0.7251,
"step": 33
},
{
"epoch": 0.16539981757373062,
"grad_norm": 0.804450273513794,
"learning_rate": 3.975903614457832e-06,
"loss": 0.731,
"step": 34
},
{
"epoch": 0.17026451809060505,
"grad_norm": 0.9799721240997314,
"learning_rate": 4.096385542168675e-06,
"loss": 0.7628,
"step": 35
},
{
"epoch": 0.1751292186074795,
"grad_norm": 0.8263784646987915,
"learning_rate": 4.216867469879519e-06,
"loss": 0.7048,
"step": 36
},
{
"epoch": 0.17999391912435392,
"grad_norm": 0.7393287420272827,
"learning_rate": 4.337349397590362e-06,
"loss": 0.7145,
"step": 37
},
{
"epoch": 0.18485861964122835,
"grad_norm": 0.6429290175437927,
"learning_rate": 4.457831325301205e-06,
"loss": 0.7006,
"step": 38
},
{
"epoch": 0.18972332015810275,
"grad_norm": 0.5760512948036194,
"learning_rate": 4.578313253012049e-06,
"loss": 0.7023,
"step": 39
},
{
"epoch": 0.19458802067497719,
"grad_norm": 0.576980471611023,
"learning_rate": 4.698795180722892e-06,
"loss": 0.7054,
"step": 40
},
{
"epoch": 0.19945272119185162,
"grad_norm": 0.5598117709159851,
"learning_rate": 4.819277108433735e-06,
"loss": 0.6959,
"step": 41
},
{
"epoch": 0.20431742170872605,
"grad_norm": 0.527066171169281,
"learning_rate": 4.939759036144578e-06,
"loss": 0.6772,
"step": 42
},
{
"epoch": 0.20918212222560048,
"grad_norm": 0.5599627494812012,
"learning_rate": 5.060240963855422e-06,
"loss": 0.6991,
"step": 43
},
{
"epoch": 0.2140468227424749,
"grad_norm": 0.5616483092308044,
"learning_rate": 5.180722891566266e-06,
"loss": 0.7013,
"step": 44
},
{
"epoch": 0.21891152325934934,
"grad_norm": 0.49973708391189575,
"learning_rate": 5.301204819277109e-06,
"loss": 0.6804,
"step": 45
},
{
"epoch": 0.22377622377622378,
"grad_norm": 0.46386387944221497,
"learning_rate": 5.421686746987952e-06,
"loss": 0.6809,
"step": 46
},
{
"epoch": 0.2286409242930982,
"grad_norm": 0.4726768732070923,
"learning_rate": 5.542168674698796e-06,
"loss": 0.676,
"step": 47
},
{
"epoch": 0.23350562480997264,
"grad_norm": 0.4434186816215515,
"learning_rate": 5.66265060240964e-06,
"loss": 0.6622,
"step": 48
},
{
"epoch": 0.23837032532684707,
"grad_norm": 0.43637368083000183,
"learning_rate": 5.783132530120482e-06,
"loss": 0.6757,
"step": 49
},
{
"epoch": 0.2432350258437215,
"grad_norm": 0.46703457832336426,
"learning_rate": 5.9036144578313255e-06,
"loss": 0.6609,
"step": 50
},
{
"epoch": 0.24809972636059593,
"grad_norm": 0.4197254180908203,
"learning_rate": 6.02409638554217e-06,
"loss": 0.6616,
"step": 51
},
{
"epoch": 0.25296442687747034,
"grad_norm": 0.4465361535549164,
"learning_rate": 6.144578313253012e-06,
"loss": 0.6956,
"step": 52
},
{
"epoch": 0.25782912739434477,
"grad_norm": 0.3944004476070404,
"learning_rate": 6.265060240963856e-06,
"loss": 0.6618,
"step": 53
},
{
"epoch": 0.2626938279112192,
"grad_norm": 0.4017980098724365,
"learning_rate": 6.385542168674699e-06,
"loss": 0.6798,
"step": 54
},
{
"epoch": 0.26755852842809363,
"grad_norm": 0.40108731389045715,
"learning_rate": 6.5060240963855425e-06,
"loss": 0.6887,
"step": 55
},
{
"epoch": 0.27242322894496807,
"grad_norm": 0.4038759768009186,
"learning_rate": 6.626506024096386e-06,
"loss": 0.6628,
"step": 56
},
{
"epoch": 0.2772879294618425,
"grad_norm": 0.40925461053848267,
"learning_rate": 6.746987951807229e-06,
"loss": 0.6542,
"step": 57
},
{
"epoch": 0.28215262997871693,
"grad_norm": 0.3734932541847229,
"learning_rate": 6.867469879518073e-06,
"loss": 0.6379,
"step": 58
},
{
"epoch": 0.28701733049559136,
"grad_norm": 0.42691197991371155,
"learning_rate": 6.987951807228917e-06,
"loss": 0.647,
"step": 59
},
{
"epoch": 0.2918820310124658,
"grad_norm": 0.4151201546192169,
"learning_rate": 7.1084337349397595e-06,
"loss": 0.6633,
"step": 60
},
{
"epoch": 0.2967467315293402,
"grad_norm": 0.3962690532207489,
"learning_rate": 7.228915662650603e-06,
"loss": 0.6629,
"step": 61
},
{
"epoch": 0.30161143204621466,
"grad_norm": 0.3834770917892456,
"learning_rate": 7.349397590361447e-06,
"loss": 0.655,
"step": 62
},
{
"epoch": 0.3064761325630891,
"grad_norm": 0.3900351822376251,
"learning_rate": 7.469879518072289e-06,
"loss": 0.632,
"step": 63
},
{
"epoch": 0.3113408330799635,
"grad_norm": 0.5288612246513367,
"learning_rate": 7.590361445783133e-06,
"loss": 0.6408,
"step": 64
},
{
"epoch": 0.31620553359683795,
"grad_norm": 0.3623826205730438,
"learning_rate": 7.710843373493977e-06,
"loss": 0.654,
"step": 65
},
{
"epoch": 0.3210702341137124,
"grad_norm": 0.3914937973022461,
"learning_rate": 7.83132530120482e-06,
"loss": 0.6428,
"step": 66
},
{
"epoch": 0.3259349346305868,
"grad_norm": 0.3521091639995575,
"learning_rate": 7.951807228915663e-06,
"loss": 0.6359,
"step": 67
},
{
"epoch": 0.33079963514746125,
"grad_norm": 0.3733057677745819,
"learning_rate": 8.072289156626508e-06,
"loss": 0.6734,
"step": 68
},
{
"epoch": 0.3356643356643357,
"grad_norm": 0.4045925438404083,
"learning_rate": 8.19277108433735e-06,
"loss": 0.6837,
"step": 69
},
{
"epoch": 0.3405290361812101,
"grad_norm": 0.3743281364440918,
"learning_rate": 8.313253012048194e-06,
"loss": 0.6448,
"step": 70
},
{
"epoch": 0.34539373669808454,
"grad_norm": 0.3639332354068756,
"learning_rate": 8.433734939759038e-06,
"loss": 0.6285,
"step": 71
},
{
"epoch": 0.350258437214959,
"grad_norm": 0.39523962140083313,
"learning_rate": 8.55421686746988e-06,
"loss": 0.6508,
"step": 72
},
{
"epoch": 0.3551231377318334,
"grad_norm": 0.3910689055919647,
"learning_rate": 8.674698795180724e-06,
"loss": 0.6378,
"step": 73
},
{
"epoch": 0.35998783824870784,
"grad_norm": 0.3988930583000183,
"learning_rate": 8.795180722891567e-06,
"loss": 0.6484,
"step": 74
},
{
"epoch": 0.36485253876558227,
"grad_norm": 0.35264673829078674,
"learning_rate": 8.91566265060241e-06,
"loss": 0.6321,
"step": 75
},
{
"epoch": 0.3697172392824567,
"grad_norm": 0.40589064359664917,
"learning_rate": 9.036144578313254e-06,
"loss": 0.6259,
"step": 76
},
{
"epoch": 0.3745819397993311,
"grad_norm": 0.35977646708488464,
"learning_rate": 9.156626506024097e-06,
"loss": 0.6456,
"step": 77
},
{
"epoch": 0.3794466403162055,
"grad_norm": 0.4060300886631012,
"learning_rate": 9.27710843373494e-06,
"loss": 0.6698,
"step": 78
},
{
"epoch": 0.38431134083307994,
"grad_norm": 0.3815556466579437,
"learning_rate": 9.397590361445785e-06,
"loss": 0.6575,
"step": 79
},
{
"epoch": 0.38917604134995437,
"grad_norm": 0.35774871706962585,
"learning_rate": 9.518072289156628e-06,
"loss": 0.6506,
"step": 80
},
{
"epoch": 0.3940407418668288,
"grad_norm": 0.4149310886859894,
"learning_rate": 9.63855421686747e-06,
"loss": 0.684,
"step": 81
},
{
"epoch": 0.39890544238370323,
"grad_norm": 0.3733583390712738,
"learning_rate": 9.759036144578315e-06,
"loss": 0.6237,
"step": 82
},
{
"epoch": 0.40377014290057767,
"grad_norm": 0.3399677574634552,
"learning_rate": 9.879518072289156e-06,
"loss": 0.6281,
"step": 83
},
{
"epoch": 0.4086348434174521,
"grad_norm": 0.37513449788093567,
"learning_rate": 1e-05,
"loss": 0.6399,
"step": 84
},
{
"epoch": 0.41349954393432653,
"grad_norm": 0.37447768449783325,
"learning_rate": 9.99995506314361e-06,
"loss": 0.633,
"step": 85
},
{
"epoch": 0.41836424445120096,
"grad_norm": 0.40050914883613586,
"learning_rate": 9.99982025338217e-06,
"loss": 0.6684,
"step": 86
},
{
"epoch": 0.4232289449680754,
"grad_norm": 0.3781862258911133,
"learning_rate": 9.999595573138845e-06,
"loss": 0.6088,
"step": 87
},
{
"epoch": 0.4280936454849498,
"grad_norm": 0.3892683982849121,
"learning_rate": 9.99928102645221e-06,
"loss": 0.6381,
"step": 88
},
{
"epoch": 0.43295834600182426,
"grad_norm": 0.37661153078079224,
"learning_rate": 9.99887661897616e-06,
"loss": 0.618,
"step": 89
},
{
"epoch": 0.4378230465186987,
"grad_norm": 0.3976646661758423,
"learning_rate": 9.99838235797981e-06,
"loss": 0.6464,
"step": 90
},
{
"epoch": 0.4426877470355731,
"grad_norm": 0.38253095746040344,
"learning_rate": 9.997798252347382e-06,
"loss": 0.5829,
"step": 91
},
{
"epoch": 0.44755244755244755,
"grad_norm": 0.4197298586368561,
"learning_rate": 9.99712431257802e-06,
"loss": 0.6105,
"step": 92
},
{
"epoch": 0.452417148069322,
"grad_norm": 0.4068802297115326,
"learning_rate": 9.996360550785619e-06,
"loss": 0.6505,
"step": 93
},
{
"epoch": 0.4572818485861964,
"grad_norm": 0.39049994945526123,
"learning_rate": 9.9955069806986e-06,
"loss": 0.6292,
"step": 94
},
{
"epoch": 0.46214654910307085,
"grad_norm": 0.41957932710647583,
"learning_rate": 9.994563617659665e-06,
"loss": 0.6304,
"step": 95
},
{
"epoch": 0.4670112496199453,
"grad_norm": 0.36546021699905396,
"learning_rate": 9.993530478625524e-06,
"loss": 0.6199,
"step": 96
},
{
"epoch": 0.4718759501368197,
"grad_norm": 0.359742134809494,
"learning_rate": 9.992407582166582e-06,
"loss": 0.594,
"step": 97
},
{
"epoch": 0.47674065065369414,
"grad_norm": 0.4413560628890991,
"learning_rate": 9.991194948466615e-06,
"loss": 0.6367,
"step": 98
},
{
"epoch": 0.4816053511705686,
"grad_norm": 0.3888995349407196,
"learning_rate": 9.989892599322404e-06,
"loss": 0.6168,
"step": 99
},
{
"epoch": 0.486470051687443,
"grad_norm": 0.41815799474716187,
"learning_rate": 9.988500558143337e-06,
"loss": 0.6445,
"step": 100
},
{
"epoch": 0.49133475220431744,
"grad_norm": 0.38159245252609253,
"learning_rate": 9.987018849950996e-06,
"loss": 0.6343,
"step": 101
},
{
"epoch": 0.49619945272119187,
"grad_norm": 0.4042263627052307,
"learning_rate": 9.985447501378706e-06,
"loss": 0.6318,
"step": 102
},
{
"epoch": 0.5010641532380663,
"grad_norm": 0.41288211941719055,
"learning_rate": 9.983786540671052e-06,
"loss": 0.6255,
"step": 103
},
{
"epoch": 0.5059288537549407,
"grad_norm": 0.40072107315063477,
"learning_rate": 9.982035997683372e-06,
"loss": 0.6213,
"step": 104
},
{
"epoch": 0.5107935542718152,
"grad_norm": 0.3677405118942261,
"learning_rate": 9.980195903881231e-06,
"loss": 0.6225,
"step": 105
},
{
"epoch": 0.5156582547886895,
"grad_norm": 0.4360586106777191,
"learning_rate": 9.978266292339838e-06,
"loss": 0.638,
"step": 106
},
{
"epoch": 0.520522955305564,
"grad_norm": 0.3555741310119629,
"learning_rate": 9.976247197743465e-06,
"loss": 0.6114,
"step": 107
},
{
"epoch": 0.5253876558224384,
"grad_norm": 0.39081236720085144,
"learning_rate": 9.974138656384815e-06,
"loss": 0.6144,
"step": 108
},
{
"epoch": 0.5302523563393129,
"grad_norm": 0.4209583103656769,
"learning_rate": 9.97194070616438e-06,
"loss": 0.6215,
"step": 109
},
{
"epoch": 0.5351170568561873,
"grad_norm": 0.3668820559978485,
"learning_rate": 9.969653386589749e-06,
"loss": 0.6264,
"step": 110
},
{
"epoch": 0.5399817573730618,
"grad_norm": 0.3862663805484772,
"learning_rate": 9.967276738774897e-06,
"loss": 0.6149,
"step": 111
},
{
"epoch": 0.5448464578899361,
"grad_norm": 0.3995136320590973,
"learning_rate": 9.964810805439464e-06,
"loss": 0.6258,
"step": 112
},
{
"epoch": 0.5497111584068106,
"grad_norm": 0.3889177143573761,
"learning_rate": 9.962255630907964e-06,
"loss": 0.6377,
"step": 113
},
{
"epoch": 0.554575858923685,
"grad_norm": 0.389662504196167,
"learning_rate": 9.959611261108999e-06,
"loss": 0.6047,
"step": 114
},
{
"epoch": 0.5594405594405595,
"grad_norm": 0.4620538651943207,
"learning_rate": 9.956877743574437e-06,
"loss": 0.644,
"step": 115
},
{
"epoch": 0.5643052599574339,
"grad_norm": 0.3749834895133972,
"learning_rate": 9.954055127438554e-06,
"loss": 0.6107,
"step": 116
},
{
"epoch": 0.5691699604743083,
"grad_norm": 0.37528854608535767,
"learning_rate": 9.951143463437145e-06,
"loss": 0.5933,
"step": 117
},
{
"epoch": 0.5740346609911827,
"grad_norm": 0.38073307275772095,
"learning_rate": 9.948142803906623e-06,
"loss": 0.6252,
"step": 118
},
{
"epoch": 0.5788993615080572,
"grad_norm": 0.4140167534351349,
"learning_rate": 9.94505320278307e-06,
"loss": 0.6329,
"step": 119
},
{
"epoch": 0.5837640620249316,
"grad_norm": 0.39784958958625793,
"learning_rate": 9.94187471560127e-06,
"loss": 0.6226,
"step": 120
},
{
"epoch": 0.5886287625418061,
"grad_norm": 0.37837842106819153,
"learning_rate": 9.938607399493714e-06,
"loss": 0.6266,
"step": 121
},
{
"epoch": 0.5934934630586804,
"grad_norm": 0.39705371856689453,
"learning_rate": 9.935251313189564e-06,
"loss": 0.6535,
"step": 122
},
{
"epoch": 0.5983581635755548,
"grad_norm": 0.39558154344558716,
"learning_rate": 9.931806517013612e-06,
"loss": 0.6261,
"step": 123
},
{
"epoch": 0.6032228640924293,
"grad_norm": 0.40359461307525635,
"learning_rate": 9.92827307288518e-06,
"loss": 0.6414,
"step": 124
},
{
"epoch": 0.6080875646093037,
"grad_norm": 0.37531161308288574,
"learning_rate": 9.924651044317017e-06,
"loss": 0.6108,
"step": 125
},
{
"epoch": 0.6129522651261782,
"grad_norm": 0.37722480297088623,
"learning_rate": 9.920940496414153e-06,
"loss": 0.6065,
"step": 126
},
{
"epoch": 0.6178169656430526,
"grad_norm": 0.3826655149459839,
"learning_rate": 9.917141495872733e-06,
"loss": 0.6189,
"step": 127
},
{
"epoch": 0.622681666159927,
"grad_norm": 0.3929085433483124,
"learning_rate": 9.913254110978812e-06,
"loss": 0.631,
"step": 128
},
{
"epoch": 0.6275463666768014,
"grad_norm": 0.3875380754470825,
"learning_rate": 9.909278411607134e-06,
"loss": 0.6211,
"step": 129
},
{
"epoch": 0.6324110671936759,
"grad_norm": 0.39089295268058777,
"learning_rate": 9.90521446921987e-06,
"loss": 0.5943,
"step": 130
},
{
"epoch": 0.6372757677105503,
"grad_norm": 0.3719761371612549,
"learning_rate": 9.90106235686534e-06,
"loss": 0.6327,
"step": 131
},
{
"epoch": 0.6421404682274248,
"grad_norm": 0.39393913745880127,
"learning_rate": 9.896822149176695e-06,
"loss": 0.636,
"step": 132
},
{
"epoch": 0.6470051687442991,
"grad_norm": 0.4049062728881836,
"learning_rate": 9.892493922370575e-06,
"loss": 0.6225,
"step": 133
},
{
"epoch": 0.6518698692611736,
"grad_norm": 0.4028378427028656,
"learning_rate": 9.888077754245741e-06,
"loss": 0.6273,
"step": 134
},
{
"epoch": 0.656734569778048,
"grad_norm": 0.41148582100868225,
"learning_rate": 9.883573724181683e-06,
"loss": 0.6459,
"step": 135
},
{
"epoch": 0.6615992702949225,
"grad_norm": 0.3935914933681488,
"learning_rate": 9.878981913137178e-06,
"loss": 0.6195,
"step": 136
},
{
"epoch": 0.6664639708117969,
"grad_norm": 0.3569656014442444,
"learning_rate": 9.87430240364885e-06,
"loss": 0.6225,
"step": 137
},
{
"epoch": 0.6713286713286714,
"grad_norm": 0.38845062255859375,
"learning_rate": 9.869535279829674e-06,
"loss": 0.62,
"step": 138
},
{
"epoch": 0.6761933718455457,
"grad_norm": 0.3890616297721863,
"learning_rate": 9.864680627367476e-06,
"loss": 0.6271,
"step": 139
},
{
"epoch": 0.6810580723624202,
"grad_norm": 0.46930286288261414,
"learning_rate": 9.859738533523384e-06,
"loss": 0.6151,
"step": 140
},
{
"epoch": 0.6859227728792946,
"grad_norm": 0.37581589818000793,
"learning_rate": 9.854709087130261e-06,
"loss": 0.6123,
"step": 141
},
{
"epoch": 0.6907874733961691,
"grad_norm": 0.4044643044471741,
"learning_rate": 9.849592378591113e-06,
"loss": 0.6426,
"step": 142
},
{
"epoch": 0.6956521739130435,
"grad_norm": 0.3923414945602417,
"learning_rate": 9.844388499877457e-06,
"loss": 0.6212,
"step": 143
},
{
"epoch": 0.700516874429918,
"grad_norm": 0.4025014042854309,
"learning_rate": 9.839097544527674e-06,
"loss": 0.6044,
"step": 144
},
{
"epoch": 0.7053815749467923,
"grad_norm": 0.3886919319629669,
"learning_rate": 9.833719607645325e-06,
"loss": 0.6019,
"step": 145
},
{
"epoch": 0.7102462754636668,
"grad_norm": 0.376140832901001,
"learning_rate": 9.82825478589744e-06,
"loss": 0.6233,
"step": 146
},
{
"epoch": 0.7151109759805412,
"grad_norm": 0.3581697642803192,
"learning_rate": 9.822703177512783e-06,
"loss": 0.6315,
"step": 147
},
{
"epoch": 0.7199756764974157,
"grad_norm": 0.34552642703056335,
"learning_rate": 9.817064882280085e-06,
"loss": 0.6063,
"step": 148
},
{
"epoch": 0.72484037701429,
"grad_norm": 0.38428688049316406,
"learning_rate": 9.811340001546252e-06,
"loss": 0.6089,
"step": 149
},
{
"epoch": 0.7297050775311645,
"grad_norm": 0.3815344274044037,
"learning_rate": 9.805528638214543e-06,
"loss": 0.6158,
"step": 150
},
{
"epoch": 0.7345697780480389,
"grad_norm": 0.37385204434394836,
"learning_rate": 9.799630896742716e-06,
"loss": 0.6148,
"step": 151
},
{
"epoch": 0.7394344785649134,
"grad_norm": 0.35990026593208313,
"learning_rate": 9.793646883141155e-06,
"loss": 0.6007,
"step": 152
},
{
"epoch": 0.7442991790817878,
"grad_norm": 0.38917872309684753,
"learning_rate": 9.787576704970965e-06,
"loss": 0.608,
"step": 153
},
{
"epoch": 0.7491638795986622,
"grad_norm": 0.36917781829833984,
"learning_rate": 9.781420471342035e-06,
"loss": 0.6177,
"step": 154
},
{
"epoch": 0.7540285801155366,
"grad_norm": 0.396499365568161,
"learning_rate": 9.77517829291108e-06,
"loss": 0.6141,
"step": 155
},
{
"epoch": 0.758893280632411,
"grad_norm": 0.39200127124786377,
"learning_rate": 9.768850281879651e-06,
"loss": 0.63,
"step": 156
},
{
"epoch": 0.7637579811492855,
"grad_norm": 0.39482221007347107,
"learning_rate": 9.762436551992117e-06,
"loss": 0.6202,
"step": 157
},
{
"epoch": 0.7686226816661599,
"grad_norm": 0.368528813123703,
"learning_rate": 9.755937218533622e-06,
"loss": 0.6111,
"step": 158
},
{
"epoch": 0.7734873821830344,
"grad_norm": 0.3803812861442566,
"learning_rate": 9.74935239832801e-06,
"loss": 0.6095,
"step": 159
},
{
"epoch": 0.7783520826999087,
"grad_norm": 0.38212645053863525,
"learning_rate": 9.742682209735727e-06,
"loss": 0.6199,
"step": 160
},
{
"epoch": 0.7832167832167832,
"grad_norm": 0.38313084840774536,
"learning_rate": 9.735926772651703e-06,
"loss": 0.6153,
"step": 161
},
{
"epoch": 0.7880814837336576,
"grad_norm": 0.3594917953014374,
"learning_rate": 9.729086208503174e-06,
"loss": 0.6264,
"step": 162
},
{
"epoch": 0.7929461842505321,
"grad_norm": 0.3762817978858948,
"learning_rate": 9.722160640247523e-06,
"loss": 0.58,
"step": 163
},
{
"epoch": 0.7978108847674065,
"grad_norm": 0.44009992480278015,
"learning_rate": 9.715150192370054e-06,
"loss": 0.6294,
"step": 164
},
{
"epoch": 0.802675585284281,
"grad_norm": 0.3405672609806061,
"learning_rate": 9.708054990881763e-06,
"loss": 0.6057,
"step": 165
},
{
"epoch": 0.8075402858011553,
"grad_norm": 0.340965211391449,
"learning_rate": 9.700875163317072e-06,
"loss": 0.6091,
"step": 166
},
{
"epoch": 0.8124049863180298,
"grad_norm": 0.375713586807251,
"learning_rate": 9.693610838731532e-06,
"loss": 0.5913,
"step": 167
},
{
"epoch": 0.8172696868349042,
"grad_norm": 0.3714539706707001,
"learning_rate": 9.686262147699507e-06,
"loss": 0.6115,
"step": 168
},
{
"epoch": 0.8221343873517787,
"grad_norm": 0.3749929666519165,
"learning_rate": 9.678829222311827e-06,
"loss": 0.6075,
"step": 169
},
{
"epoch": 0.8269990878686531,
"grad_norm": 0.3718150556087494,
"learning_rate": 9.671312196173413e-06,
"loss": 0.6178,
"step": 170
},
{
"epoch": 0.8318637883855275,
"grad_norm": 0.4176604449748993,
"learning_rate": 9.663711204400872e-06,
"loss": 0.6085,
"step": 171
},
{
"epoch": 0.8367284889024019,
"grad_norm": 0.422677218914032,
"learning_rate": 9.656026383620076e-06,
"loss": 0.6242,
"step": 172
},
{
"epoch": 0.8415931894192764,
"grad_norm": 0.372397780418396,
"learning_rate": 9.6482578719637e-06,
"loss": 0.6259,
"step": 173
},
{
"epoch": 0.8464578899361508,
"grad_norm": 0.42137330770492554,
"learning_rate": 9.640405809068743e-06,
"loss": 0.6223,
"step": 174
},
{
"epoch": 0.8513225904530253,
"grad_norm": 0.39700597524642944,
"learning_rate": 9.632470336074009e-06,
"loss": 0.6146,
"step": 175
},
{
"epoch": 0.8561872909698997,
"grad_norm": 0.3760301172733307,
"learning_rate": 9.624451595617588e-06,
"loss": 0.6095,
"step": 176
},
{
"epoch": 0.8610519914867741,
"grad_norm": 0.4247523546218872,
"learning_rate": 9.616349731834271e-06,
"loss": 0.6215,
"step": 177
},
{
"epoch": 0.8659166920036485,
"grad_norm": 0.39961710572242737,
"learning_rate": 9.608164890352977e-06,
"loss": 0.641,
"step": 178
},
{
"epoch": 0.870781392520523,
"grad_norm": 0.3871230185031891,
"learning_rate": 9.599897218294122e-06,
"loss": 0.594,
"step": 179
},
{
"epoch": 0.8756460930373974,
"grad_norm": 0.3941272795200348,
"learning_rate": 9.591546864266983e-06,
"loss": 0.5826,
"step": 180
},
{
"epoch": 0.8805107935542719,
"grad_norm": 0.3947692811489105,
"learning_rate": 9.583113978367026e-06,
"loss": 0.6061,
"step": 181
},
{
"epoch": 0.8853754940711462,
"grad_norm": 0.39284199476242065,
"learning_rate": 9.574598712173202e-06,
"loss": 0.6175,
"step": 182
},
{
"epoch": 0.8902401945880207,
"grad_norm": 0.41883713006973267,
"learning_rate": 9.56600121874523e-06,
"loss": 0.6124,
"step": 183
},
{
"epoch": 0.8951048951048951,
"grad_norm": 0.38454899191856384,
"learning_rate": 9.557321652620839e-06,
"loss": 0.6213,
"step": 184
},
{
"epoch": 0.8999695956217695,
"grad_norm": 0.38847121596336365,
"learning_rate": 9.548560169812997e-06,
"loss": 0.6178,
"step": 185
},
{
"epoch": 0.904834296138644,
"grad_norm": 0.4160056412220001,
"learning_rate": 9.539716927807102e-06,
"loss": 0.6176,
"step": 186
},
{
"epoch": 0.9096989966555183,
"grad_norm": 0.36343953013420105,
"learning_rate": 9.530792085558151e-06,
"loss": 0.604,
"step": 187
},
{
"epoch": 0.9145636971723928,
"grad_norm": 0.40518003702163696,
"learning_rate": 9.521785803487888e-06,
"loss": 0.6212,
"step": 188
},
{
"epoch": 0.9194283976892672,
"grad_norm": 0.39145195484161377,
"learning_rate": 9.512698243481914e-06,
"loss": 0.6281,
"step": 189
},
{
"epoch": 0.9242930982061417,
"grad_norm": 0.39723438024520874,
"learning_rate": 9.50352956888678e-06,
"loss": 0.6128,
"step": 190
},
{
"epoch": 0.9291577987230161,
"grad_norm": 0.3605063855648041,
"learning_rate": 9.49427994450705e-06,
"loss": 0.5888,
"step": 191
},
{
"epoch": 0.9340224992398906,
"grad_norm": 0.3381179869174957,
"learning_rate": 9.484949536602343e-06,
"loss": 0.6191,
"step": 192
},
{
"epoch": 0.9388871997567649,
"grad_norm": 0.3913276195526123,
"learning_rate": 9.47553851288434e-06,
"loss": 0.612,
"step": 193
},
{
"epoch": 0.9437519002736394,
"grad_norm": 0.3754216432571411,
"learning_rate": 9.466047042513767e-06,
"loss": 0.6239,
"step": 194
},
{
"epoch": 0.9486166007905138,
"grad_norm": 0.35542696714401245,
"learning_rate": 9.45647529609736e-06,
"loss": 0.6173,
"step": 195
},
{
"epoch": 0.9534813013073883,
"grad_norm": 0.34806758165359497,
"learning_rate": 9.4468234456848e-06,
"loss": 0.6279,
"step": 196
},
{
"epoch": 0.9583460018242627,
"grad_norm": 0.34163591265678406,
"learning_rate": 9.437091664765611e-06,
"loss": 0.5814,
"step": 197
},
{
"epoch": 0.9632107023411371,
"grad_norm": 0.3506777882575989,
"learning_rate": 9.427280128266049e-06,
"loss": 0.608,
"step": 198
},
{
"epoch": 0.9680754028580115,
"grad_norm": 0.39640653133392334,
"learning_rate": 9.41738901254596e-06,
"loss": 0.6242,
"step": 199
},
{
"epoch": 0.972940103374886,
"grad_norm": 0.38849350810050964,
"learning_rate": 9.4074184953956e-06,
"loss": 0.6001,
"step": 200
},
{
"epoch": 0.9778048038917604,
"grad_norm": 0.34266453981399536,
"learning_rate": 9.397368756032445e-06,
"loss": 0.6072,
"step": 201
},
{
"epoch": 0.9826695044086349,
"grad_norm": 0.3617665767669678,
"learning_rate": 9.38723997509798e-06,
"loss": 0.6072,
"step": 202
},
{
"epoch": 0.9875342049255093,
"grad_norm": 0.37782296538352966,
"learning_rate": 9.37703233465443e-06,
"loss": 0.6123,
"step": 203
},
{
"epoch": 0.9923989054423837,
"grad_norm": 0.392240047454834,
"learning_rate": 9.366746018181503e-06,
"loss": 0.6202,
"step": 204
},
{
"epoch": 0.9972636059592581,
"grad_norm": 0.3623761832714081,
"learning_rate": 9.356381210573092e-06,
"loss": 0.5883,
"step": 205
},
{
"epoch": 1.0,
"grad_norm": 0.5044242739677429,
"learning_rate": 9.345938098133946e-06,
"loss": 0.5917,
"step": 206
},
{
"epoch": 1.0048647005168745,
"grad_norm": 0.3975219428539276,
"learning_rate": 9.33541686857632e-06,
"loss": 0.5713,
"step": 207
},
{
"epoch": 1.0097294010337488,
"grad_norm": 0.43212786316871643,
"learning_rate": 9.324817711016609e-06,
"loss": 0.5864,
"step": 208
},
{
"epoch": 1.0145941015506232,
"grad_norm": 0.36350852251052856,
"learning_rate": 9.31414081597194e-06,
"loss": 0.5826,
"step": 209
},
{
"epoch": 1.0194588020674977,
"grad_norm": 0.38387811183929443,
"learning_rate": 9.303386375356752e-06,
"loss": 0.5715,
"step": 210
},
{
"epoch": 1.0243235025843722,
"grad_norm": 0.39839062094688416,
"learning_rate": 9.292554582479349e-06,
"loss": 0.5839,
"step": 211
},
{
"epoch": 1.0291882031012465,
"grad_norm": 0.6260845065116882,
"learning_rate": 9.281645632038417e-06,
"loss": 0.6068,
"step": 212
},
{
"epoch": 1.034052903618121,
"grad_norm": 0.3589637577533722,
"learning_rate": 9.270659720119533e-06,
"loss": 0.577,
"step": 213
},
{
"epoch": 1.0389176041349955,
"grad_norm": 0.38663890957832336,
"learning_rate": 9.259597044191635e-06,
"loss": 0.5815,
"step": 214
},
{
"epoch": 1.04378230465187,
"grad_norm": 0.4027801752090454,
"learning_rate": 9.248457803103476e-06,
"loss": 0.5445,
"step": 215
},
{
"epoch": 1.0486470051687442,
"grad_norm": 0.4009842574596405,
"learning_rate": 9.237242197080045e-06,
"loss": 0.5799,
"step": 216
},
{
"epoch": 1.0535117056856187,
"grad_norm": 0.40365034341812134,
"learning_rate": 9.225950427718974e-06,
"loss": 0.6012,
"step": 217
},
{
"epoch": 1.0583764062024932,
"grad_norm": 0.3974784016609192,
"learning_rate": 9.21458269798691e-06,
"loss": 0.5915,
"step": 218
},
{
"epoch": 1.0632411067193677,
"grad_norm": 0.45469483733177185,
"learning_rate": 9.203139212215868e-06,
"loss": 0.6232,
"step": 219
},
{
"epoch": 1.068105807236242,
"grad_norm": 0.4087798297405243,
"learning_rate": 9.191620176099559e-06,
"loss": 0.6025,
"step": 220
},
{
"epoch": 1.0729705077531164,
"grad_norm": 0.37921905517578125,
"learning_rate": 9.180025796689692e-06,
"loss": 0.5898,
"step": 221
},
{
"epoch": 1.077835208269991,
"grad_norm": 0.41409194469451904,
"learning_rate": 9.168356282392253e-06,
"loss": 0.5812,
"step": 222
},
{
"epoch": 1.0826999087868654,
"grad_norm": 1.3059632778167725,
"learning_rate": 9.156611842963753e-06,
"loss": 0.58,
"step": 223
},
{
"epoch": 1.0875646093037397,
"grad_norm": 0.42920663952827454,
"learning_rate": 9.144792689507471e-06,
"loss": 0.5705,
"step": 224
},
{
"epoch": 1.0924293098206141,
"grad_norm": 0.38777294754981995,
"learning_rate": 9.132899034469648e-06,
"loss": 0.5853,
"step": 225
},
{
"epoch": 1.0972940103374886,
"grad_norm": 0.41427531838417053,
"learning_rate": 9.120931091635669e-06,
"loss": 0.6141,
"step": 226
},
{
"epoch": 1.1021587108543631,
"grad_norm": 0.3924216330051422,
"learning_rate": 9.108889076126226e-06,
"loss": 0.5783,
"step": 227
},
{
"epoch": 1.1070234113712374,
"grad_norm": 0.3683719038963318,
"learning_rate": 9.09677320439345e-06,
"loss": 0.5957,
"step": 228
},
{
"epoch": 1.1118881118881119,
"grad_norm": 0.3637790381908417,
"learning_rate": 9.084583694217012e-06,
"loss": 0.558,
"step": 229
},
{
"epoch": 1.1167528124049864,
"grad_norm": 0.36177799105644226,
"learning_rate": 9.072320764700223e-06,
"loss": 0.5802,
"step": 230
},
{
"epoch": 1.1216175129218608,
"grad_norm": 0.3491053879261017,
"learning_rate": 9.059984636266082e-06,
"loss": 0.5684,
"step": 231
},
{
"epoch": 1.1264822134387351,
"grad_norm": 0.3962339162826538,
"learning_rate": 9.047575530653324e-06,
"loss": 0.6006,
"step": 232
},
{
"epoch": 1.1313469139556096,
"grad_norm": 0.3437581956386566,
"learning_rate": 9.035093670912424e-06,
"loss": 0.5806,
"step": 233
},
{
"epoch": 1.136211614472484,
"grad_norm": 0.36129409074783325,
"learning_rate": 9.022539281401601e-06,
"loss": 0.5935,
"step": 234
},
{
"epoch": 1.1410763149893586,
"grad_norm": 0.3915378749370575,
"learning_rate": 9.009912587782772e-06,
"loss": 0.5917,
"step": 235
},
{
"epoch": 1.1459410155062328,
"grad_norm": 0.35055333375930786,
"learning_rate": 8.997213817017508e-06,
"loss": 0.6044,
"step": 236
},
{
"epoch": 1.1508057160231073,
"grad_norm": 0.32850873470306396,
"learning_rate": 8.984443197362938e-06,
"loss": 0.5585,
"step": 237
},
{
"epoch": 1.1556704165399818,
"grad_norm": 0.3512062132358551,
"learning_rate": 8.971600958367668e-06,
"loss": 0.5739,
"step": 238
},
{
"epoch": 1.160535117056856,
"grad_norm": 0.3887694776058197,
"learning_rate": 8.958687330867634e-06,
"loss": 0.583,
"step": 239
},
{
"epoch": 1.1653998175737306,
"grad_norm": 0.31317535042762756,
"learning_rate": 8.94570254698197e-06,
"loss": 0.5753,
"step": 240
},
{
"epoch": 1.170264518090605,
"grad_norm": 0.36693888902664185,
"learning_rate": 8.932646840108818e-06,
"loss": 0.589,
"step": 241
},
{
"epoch": 1.1751292186074795,
"grad_norm": 0.35755452513694763,
"learning_rate": 8.919520444921153e-06,
"loss": 0.5888,
"step": 242
},
{
"epoch": 1.179993919124354,
"grad_norm": 0.3393310010433197,
"learning_rate": 8.906323597362547e-06,
"loss": 0.5586,
"step": 243
},
{
"epoch": 1.1848586196412283,
"grad_norm": 0.33648696541786194,
"learning_rate": 8.893056534642938e-06,
"loss": 0.5388,
"step": 244
},
{
"epoch": 1.1897233201581028,
"grad_norm": 0.636298656463623,
"learning_rate": 8.879719495234363e-06,
"loss": 0.5905,
"step": 245
},
{
"epoch": 1.1945880206749773,
"grad_norm": 0.3570788502693176,
"learning_rate": 8.866312718866669e-06,
"loss": 0.5756,
"step": 246
},
{
"epoch": 1.1994527211918515,
"grad_norm": 0.36703434586524963,
"learning_rate": 8.852836446523213e-06,
"loss": 0.5713,
"step": 247
},
{
"epoch": 1.204317421708726,
"grad_norm": 0.37416139245033264,
"learning_rate": 8.83929092043652e-06,
"loss": 0.5755,
"step": 248
},
{
"epoch": 1.2091821222256005,
"grad_norm": 0.356198787689209,
"learning_rate": 8.825676384083936e-06,
"loss": 0.5679,
"step": 249
},
{
"epoch": 1.214046822742475,
"grad_norm": 0.3884102404117584,
"learning_rate": 8.811993082183243e-06,
"loss": 0.5661,
"step": 250
},
{
"epoch": 1.2189115232593493,
"grad_norm": 0.3469521701335907,
"learning_rate": 8.798241260688273e-06,
"loss": 0.5944,
"step": 251
},
{
"epoch": 1.2237762237762237,
"grad_norm": 0.3841468095779419,
"learning_rate": 8.784421166784476e-06,
"loss": 0.5955,
"step": 252
},
{
"epoch": 1.2286409242930982,
"grad_norm": 0.336848646402359,
"learning_rate": 8.770533048884483e-06,
"loss": 0.5749,
"step": 253
},
{
"epoch": 1.2335056248099727,
"grad_norm": 0.38999801874160767,
"learning_rate": 8.756577156623636e-06,
"loss": 0.5947,
"step": 254
},
{
"epoch": 1.238370325326847,
"grad_norm": 0.366953581571579,
"learning_rate": 8.742553740855507e-06,
"loss": 0.561,
"step": 255
},
{
"epoch": 1.2432350258437215,
"grad_norm": 0.4165610671043396,
"learning_rate": 8.728463053647382e-06,
"loss": 0.5822,
"step": 256
},
{
"epoch": 1.248099726360596,
"grad_norm": 0.3741051256656647,
"learning_rate": 8.71430534827574e-06,
"loss": 0.5916,
"step": 257
},
{
"epoch": 1.2529644268774702,
"grad_norm": 0.3687206208705902,
"learning_rate": 8.700080879221689e-06,
"loss": 0.573,
"step": 258
},
{
"epoch": 1.2578291273943447,
"grad_norm": 0.39223089814186096,
"learning_rate": 8.685789902166395e-06,
"loss": 0.6035,
"step": 259
},
{
"epoch": 1.2626938279112192,
"grad_norm": 0.3949979841709137,
"learning_rate": 8.671432673986493e-06,
"loss": 0.5726,
"step": 260
},
{
"epoch": 1.2675585284280937,
"grad_norm": 0.36847877502441406,
"learning_rate": 8.657009452749466e-06,
"loss": 0.5909,
"step": 261
},
{
"epoch": 1.2724232289449682,
"grad_norm": 0.35309484601020813,
"learning_rate": 8.642520497709001e-06,
"loss": 0.5796,
"step": 262
},
{
"epoch": 1.2772879294618424,
"grad_norm": 0.37753936648368835,
"learning_rate": 8.627966069300332e-06,
"loss": 0.5959,
"step": 263
},
{
"epoch": 1.282152629978717,
"grad_norm": 0.3944050371646881,
"learning_rate": 8.613346429135567e-06,
"loss": 0.5806,
"step": 264
},
{
"epoch": 1.2870173304955914,
"grad_norm": 0.3692607581615448,
"learning_rate": 8.598661839998972e-06,
"loss": 0.5724,
"step": 265
},
{
"epoch": 1.2918820310124657,
"grad_norm": 0.39939919114112854,
"learning_rate": 8.583912565842258e-06,
"loss": 0.6082,
"step": 266
},
{
"epoch": 1.2967467315293402,
"grad_norm": 0.3562261164188385,
"learning_rate": 8.569098871779828e-06,
"loss": 0.566,
"step": 267
},
{
"epoch": 1.3016114320462147,
"grad_norm": 0.3693968951702118,
"learning_rate": 8.554221024084019e-06,
"loss": 0.5681,
"step": 268
},
{
"epoch": 1.3064761325630891,
"grad_norm": 0.4063265919685364,
"learning_rate": 8.539279290180315e-06,
"loss": 0.5987,
"step": 269
},
{
"epoch": 1.3113408330799636,
"grad_norm": 0.3675994277000427,
"learning_rate": 8.524273938642539e-06,
"loss": 0.5833,
"step": 270
},
{
"epoch": 1.316205533596838,
"grad_norm": 0.35409554839134216,
"learning_rate": 8.509205239188017e-06,
"loss": 0.5908,
"step": 271
},
{
"epoch": 1.3210702341137124,
"grad_norm": 0.3764541447162628,
"learning_rate": 8.494073462672743e-06,
"loss": 0.5922,
"step": 272
},
{
"epoch": 1.3259349346305869,
"grad_norm": 0.37606513500213623,
"learning_rate": 8.478878881086505e-06,
"loss": 0.6006,
"step": 273
},
{
"epoch": 1.3307996351474611,
"grad_norm": 0.3864189088344574,
"learning_rate": 8.463621767547998e-06,
"loss": 0.5959,
"step": 274
},
{
"epoch": 1.3356643356643356,
"grad_norm": 0.3626885712146759,
"learning_rate": 8.448302396299906e-06,
"loss": 0.5952,
"step": 275
},
{
"epoch": 1.34052903618121,
"grad_norm": 0.3946470022201538,
"learning_rate": 8.432921042703985e-06,
"loss": 0.5691,
"step": 276
},
{
"epoch": 1.3453937366980846,
"grad_norm": 0.3896423876285553,
"learning_rate": 8.417477983236107e-06,
"loss": 0.5899,
"step": 277
},
{
"epoch": 1.350258437214959,
"grad_norm": 0.3613225221633911,
"learning_rate": 8.401973495481289e-06,
"loss": 0.6184,
"step": 278
},
{
"epoch": 1.3551231377318333,
"grad_norm": 0.3720517158508301,
"learning_rate": 8.386407858128707e-06,
"loss": 0.5954,
"step": 279
},
{
"epoch": 1.3599878382487078,
"grad_norm": 0.3605566918849945,
"learning_rate": 8.370781350966683e-06,
"loss": 0.5659,
"step": 280
},
{
"epoch": 1.3648525387655823,
"grad_norm": 0.3908631205558777,
"learning_rate": 8.355094254877665e-06,
"loss": 0.5848,
"step": 281
},
{
"epoch": 1.3697172392824566,
"grad_norm": 0.3583967089653015,
"learning_rate": 8.339346851833163e-06,
"loss": 0.6031,
"step": 282
},
{
"epoch": 1.374581939799331,
"grad_norm": 0.3706776201725006,
"learning_rate": 8.323539424888695e-06,
"loss": 0.5906,
"step": 283
},
{
"epoch": 1.3794466403162056,
"grad_norm": 0.4411206841468811,
"learning_rate": 8.30767225817869e-06,
"loss": 0.6058,
"step": 284
},
{
"epoch": 1.3843113408330798,
"grad_norm": 0.3731936812400818,
"learning_rate": 8.291745636911382e-06,
"loss": 0.6014,
"step": 285
},
{
"epoch": 1.3891760413499543,
"grad_norm": 0.3520592153072357,
"learning_rate": 8.27575984736369e-06,
"loss": 0.5757,
"step": 286
},
{
"epoch": 1.3940407418668288,
"grad_norm": 0.3719932734966278,
"learning_rate": 8.259715176876069e-06,
"loss": 0.5834,
"step": 287
},
{
"epoch": 1.3989054423837033,
"grad_norm": 0.3510180413722992,
"learning_rate": 8.243611913847337e-06,
"loss": 0.5846,
"step": 288
},
{
"epoch": 1.4037701429005778,
"grad_norm": 0.3617526590824127,
"learning_rate": 8.2274503477295e-06,
"loss": 0.5578,
"step": 289
},
{
"epoch": 1.408634843417452,
"grad_norm": 0.36835089325904846,
"learning_rate": 8.211230769022552e-06,
"loss": 0.5726,
"step": 290
},
{
"epoch": 1.4134995439343265,
"grad_norm": 0.40028098225593567,
"learning_rate": 8.19495346926924e-06,
"loss": 0.6066,
"step": 291
},
{
"epoch": 1.418364244451201,
"grad_norm": 0.39484137296676636,
"learning_rate": 8.178618741049841e-06,
"loss": 0.5958,
"step": 292
},
{
"epoch": 1.4232289449680753,
"grad_norm": 0.32314619421958923,
"learning_rate": 8.162226877976886e-06,
"loss": 0.5486,
"step": 293
},
{
"epoch": 1.4280936454849498,
"grad_norm": 0.3897193372249603,
"learning_rate": 8.145778174689897e-06,
"loss": 0.604,
"step": 294
},
{
"epoch": 1.4329583460018243,
"grad_norm": 0.3503423035144806,
"learning_rate": 8.129272926850079e-06,
"loss": 0.5767,
"step": 295
},
{
"epoch": 1.4378230465186987,
"grad_norm": 0.39865437150001526,
"learning_rate": 8.112711431135014e-06,
"loss": 0.5996,
"step": 296
},
{
"epoch": 1.4426877470355732,
"grad_norm": 0.34130704402923584,
"learning_rate": 8.096093985233323e-06,
"loss": 0.5751,
"step": 297
},
{
"epoch": 1.4475524475524475,
"grad_norm": 0.37699124217033386,
"learning_rate": 8.079420887839316e-06,
"loss": 0.565,
"step": 298
},
{
"epoch": 1.452417148069322,
"grad_norm": 0.37764230370521545,
"learning_rate": 8.062692438647628e-06,
"loss": 0.599,
"step": 299
},
{
"epoch": 1.4572818485861965,
"grad_norm": 0.3869394361972809,
"learning_rate": 8.045908938347828e-06,
"loss": 0.5533,
"step": 300
},
{
"epoch": 1.4621465491030707,
"grad_norm": 0.33982014656066895,
"learning_rate": 8.029070688619013e-06,
"loss": 0.5482,
"step": 301
},
{
"epoch": 1.4670112496199452,
"grad_norm": 0.38899099826812744,
"learning_rate": 8.012177992124385e-06,
"loss": 0.5688,
"step": 302
},
{
"epoch": 1.4718759501368197,
"grad_norm": 0.3462952971458435,
"learning_rate": 7.995231152505815e-06,
"loss": 0.5847,
"step": 303
},
{
"epoch": 1.4767406506536942,
"grad_norm": 0.4081119894981384,
"learning_rate": 7.978230474378383e-06,
"loss": 0.5869,
"step": 304
},
{
"epoch": 1.4816053511705687,
"grad_norm": 0.39347121119499207,
"learning_rate": 7.961176263324902e-06,
"loss": 0.6087,
"step": 305
},
{
"epoch": 1.486470051687443,
"grad_norm": 0.4140094518661499,
"learning_rate": 7.944068825890424e-06,
"loss": 0.6283,
"step": 306
},
{
"epoch": 1.4913347522043174,
"grad_norm": 0.3770853877067566,
"learning_rate": 7.92690846957673e-06,
"loss": 0.5891,
"step": 307
},
{
"epoch": 1.496199452721192,
"grad_norm": 0.37083595991134644,
"learning_rate": 7.909695502836814e-06,
"loss": 0.5913,
"step": 308
},
{
"epoch": 1.5010641532380662,
"grad_norm": 0.3645496666431427,
"learning_rate": 7.892430235069317e-06,
"loss": 0.5855,
"step": 309
},
{
"epoch": 1.5059288537549407,
"grad_norm": 0.38950780034065247,
"learning_rate": 7.875112976612984e-06,
"loss": 0.5966,
"step": 310
},
{
"epoch": 1.5107935542718152,
"grad_norm": 0.34090083837509155,
"learning_rate": 7.857744038741076e-06,
"loss": 0.5893,
"step": 311
},
{
"epoch": 1.5156582547886894,
"grad_norm": 0.39991140365600586,
"learning_rate": 7.84032373365578e-06,
"loss": 0.6067,
"step": 312
},
{
"epoch": 1.5205229553055641,
"grad_norm": 0.34889593720436096,
"learning_rate": 7.822852374482597e-06,
"loss": 0.5726,
"step": 313
},
{
"epoch": 1.5253876558224384,
"grad_norm": 0.37084507942199707,
"learning_rate": 7.805330275264707e-06,
"loss": 0.6003,
"step": 314
},
{
"epoch": 1.530252356339313,
"grad_norm": 0.3561820387840271,
"learning_rate": 7.787757750957335e-06,
"loss": 0.5622,
"step": 315
},
{
"epoch": 1.5351170568561874,
"grad_norm": 0.3784889876842499,
"learning_rate": 7.77013511742208e-06,
"loss": 0.5807,
"step": 316
},
{
"epoch": 1.5399817573730616,
"grad_norm": 0.3501896858215332,
"learning_rate": 7.752462691421245e-06,
"loss": 0.589,
"step": 317
},
{
"epoch": 1.5448464578899361,
"grad_norm": 0.39493295550346375,
"learning_rate": 7.734740790612137e-06,
"loss": 0.5787,
"step": 318
},
{
"epoch": 1.5497111584068106,
"grad_norm": 0.3801139295101166,
"learning_rate": 7.716969733541357e-06,
"loss": 0.5951,
"step": 319
},
{
"epoch": 1.5545758589236849,
"grad_norm": 0.3728678524494171,
"learning_rate": 7.699149839639086e-06,
"loss": 0.5844,
"step": 320
},
{
"epoch": 1.5594405594405596,
"grad_norm": 0.4056283235549927,
"learning_rate": 7.681281429213328e-06,
"loss": 0.5865,
"step": 321
},
{
"epoch": 1.5643052599574339,
"grad_norm": 0.3417642414569855,
"learning_rate": 7.663364823444157e-06,
"loss": 0.5671,
"step": 322
},
{
"epoch": 1.5691699604743083,
"grad_norm": 0.3736799359321594,
"learning_rate": 7.645400344377953e-06,
"loss": 0.5868,
"step": 323
},
{
"epoch": 1.5740346609911828,
"grad_norm": 0.36077746748924255,
"learning_rate": 7.627388314921602e-06,
"loss": 0.5914,
"step": 324
},
{
"epoch": 1.578899361508057,
"grad_norm": 0.419326514005661,
"learning_rate": 7.609329058836694e-06,
"loss": 0.6124,
"step": 325
},
{
"epoch": 1.5837640620249316,
"grad_norm": 0.36732929944992065,
"learning_rate": 7.59122290073371e-06,
"loss": 0.5705,
"step": 326
},
{
"epoch": 1.588628762541806,
"grad_norm": 0.3390752673149109,
"learning_rate": 7.5730701660661795e-06,
"loss": 0.5678,
"step": 327
},
{
"epoch": 1.5934934630586803,
"grad_norm": 0.36172622442245483,
"learning_rate": 7.554871181124836e-06,
"loss": 0.5489,
"step": 328
},
{
"epoch": 1.5983581635755548,
"grad_norm": 0.37669700384140015,
"learning_rate": 7.536626273031747e-06,
"loss": 0.5952,
"step": 329
},
{
"epoch": 1.6032228640924293,
"grad_norm": 0.37252306938171387,
"learning_rate": 7.5183357697344395e-06,
"loss": 0.5841,
"step": 330
},
{
"epoch": 1.6080875646093036,
"grad_norm": 0.365755170583725,
"learning_rate": 7.500000000000001e-06,
"loss": 0.5985,
"step": 331
},
{
"epoch": 1.6129522651261783,
"grad_norm": 0.42082270979881287,
"learning_rate": 7.481619293409173e-06,
"loss": 0.583,
"step": 332
},
{
"epoch": 1.6178169656430526,
"grad_norm": 0.3403206765651703,
"learning_rate": 7.4631939803504215e-06,
"loss": 0.5448,
"step": 333
},
{
"epoch": 1.622681666159927,
"grad_norm": 0.39089831709861755,
"learning_rate": 7.44472439201401e-06,
"loss": 0.5815,
"step": 334
},
{
"epoch": 1.6275463666768015,
"grad_norm": 0.354023277759552,
"learning_rate": 7.426210860386032e-06,
"loss": 0.5777,
"step": 335
},
{
"epoch": 1.6324110671936758,
"grad_norm": 0.33848828077316284,
"learning_rate": 7.407653718242449e-06,
"loss": 0.5328,
"step": 336
},
{
"epoch": 1.6372757677105503,
"grad_norm": 0.35336801409721375,
"learning_rate": 7.3890532991431174e-06,
"loss": 0.5729,
"step": 337
},
{
"epoch": 1.6421404682274248,
"grad_norm": 0.359391450881958,
"learning_rate": 7.370409937425781e-06,
"loss": 0.5598,
"step": 338
},
{
"epoch": 1.647005168744299,
"grad_norm": 0.365854412317276,
"learning_rate": 7.3517239682000675e-06,
"loss": 0.573,
"step": 339
},
{
"epoch": 1.6518698692611737,
"grad_norm": 0.3494594693183899,
"learning_rate": 7.332995727341462e-06,
"loss": 0.5983,
"step": 340
},
{
"epoch": 1.656734569778048,
"grad_norm": 0.3526364266872406,
"learning_rate": 7.314225551485273e-06,
"loss": 0.5843,
"step": 341
},
{
"epoch": 1.6615992702949225,
"grad_norm": 0.35377034544944763,
"learning_rate": 7.295413778020579e-06,
"loss": 0.5607,
"step": 342
},
{
"epoch": 1.666463970811797,
"grad_norm": 0.3913540542125702,
"learning_rate": 7.276560745084167e-06,
"loss": 0.5937,
"step": 343
},
{
"epoch": 1.6713286713286712,
"grad_norm": 0.40563780069351196,
"learning_rate": 7.257666791554448e-06,
"loss": 0.6006,
"step": 344
},
{
"epoch": 1.6761933718455457,
"grad_norm": 0.3545846939086914,
"learning_rate": 7.2387322570453724e-06,
"loss": 0.5669,
"step": 345
},
{
"epoch": 1.6810580723624202,
"grad_norm": 0.34502899646759033,
"learning_rate": 7.219757481900325e-06,
"loss": 0.568,
"step": 346
},
{
"epoch": 1.6859227728792945,
"grad_norm": 0.39381837844848633,
"learning_rate": 7.2007428071860045e-06,
"loss": 0.588,
"step": 347
},
{
"epoch": 1.6907874733961692,
"grad_norm": 0.3615321218967438,
"learning_rate": 7.181688574686292e-06,
"loss": 0.5558,
"step": 348
},
{
"epoch": 1.6956521739130435,
"grad_norm": 0.39073994755744934,
"learning_rate": 7.162595126896111e-06,
"loss": 0.5913,
"step": 349
},
{
"epoch": 1.700516874429918,
"grad_norm": 0.38636699318885803,
"learning_rate": 7.143462807015271e-06,
"loss": 0.5892,
"step": 350
},
{
"epoch": 1.7053815749467924,
"grad_norm": 0.412045955657959,
"learning_rate": 7.1242919589422974e-06,
"loss": 0.5836,
"step": 351
},
{
"epoch": 1.7102462754636667,
"grad_norm": 0.3851414620876312,
"learning_rate": 7.105082927268247e-06,
"loss": 0.5747,
"step": 352
},
{
"epoch": 1.7151109759805412,
"grad_norm": 0.4149802625179291,
"learning_rate": 7.085836057270521e-06,
"loss": 0.5844,
"step": 353
},
{
"epoch": 1.7199756764974157,
"grad_norm": 0.37356290221214294,
"learning_rate": 7.066551694906651e-06,
"loss": 0.5703,
"step": 354
},
{
"epoch": 1.72484037701429,
"grad_norm": 0.3996489346027374,
"learning_rate": 7.047230186808085e-06,
"loss": 0.5882,
"step": 355
},
{
"epoch": 1.7297050775311646,
"grad_norm": 0.3939168453216553,
"learning_rate": 7.027871880273959e-06,
"loss": 0.5767,
"step": 356
},
{
"epoch": 1.734569778048039,
"grad_norm": 0.36278340220451355,
"learning_rate": 7.008477123264849e-06,
"loss": 0.5744,
"step": 357
},
{
"epoch": 1.7394344785649134,
"grad_norm": 0.36954575777053833,
"learning_rate": 6.989046264396516e-06,
"loss": 0.5803,
"step": 358
},
{
"epoch": 1.7442991790817879,
"grad_norm": 0.36300379037857056,
"learning_rate": 6.96957965293365e-06,
"loss": 0.5768,
"step": 359
},
{
"epoch": 1.7491638795986622,
"grad_norm": 0.3576260209083557,
"learning_rate": 6.9500776387835785e-06,
"loss": 0.5518,
"step": 360
},
{
"epoch": 1.7540285801155366,
"grad_norm": 0.36793169379234314,
"learning_rate": 6.9305405724899876e-06,
"loss": 0.585,
"step": 361
},
{
"epoch": 1.7588932806324111,
"grad_norm": 0.3626561164855957,
"learning_rate": 6.91096880522661e-06,
"loss": 0.6082,
"step": 362
},
{
"epoch": 1.7637579811492854,
"grad_norm": 0.3631689250469208,
"learning_rate": 6.891362688790925e-06,
"loss": 0.5907,
"step": 363
},
{
"epoch": 1.7686226816661599,
"grad_norm": 0.36881235241889954,
"learning_rate": 6.871722575597829e-06,
"loss": 0.5592,
"step": 364
},
{
"epoch": 1.7734873821830344,
"grad_norm": 0.31317824125289917,
"learning_rate": 6.8520488186733e-06,
"loss": 0.5631,
"step": 365
},
{
"epoch": 1.7783520826999086,
"grad_norm": 0.3747887909412384,
"learning_rate": 6.832341771648057e-06,
"loss": 0.6121,
"step": 366
},
{
"epoch": 1.7832167832167833,
"grad_norm": 0.3274476230144501,
"learning_rate": 6.812601788751192e-06,
"loss": 0.5809,
"step": 367
},
{
"epoch": 1.7880814837336576,
"grad_norm": 0.31376752257347107,
"learning_rate": 6.792829224803816e-06,
"loss": 0.576,
"step": 368
},
{
"epoch": 1.792946184250532,
"grad_norm": 0.34662845730781555,
"learning_rate": 6.773024435212678e-06,
"loss": 0.5831,
"step": 369
},
{
"epoch": 1.7978108847674066,
"grad_norm": 0.39459121227264404,
"learning_rate": 6.753187775963773e-06,
"loss": 0.6016,
"step": 370
},
{
"epoch": 1.8026755852842808,
"grad_norm": 0.34671536087989807,
"learning_rate": 6.733319603615941e-06,
"loss": 0.5568,
"step": 371
},
{
"epoch": 1.8075402858011553,
"grad_norm": 0.3303484618663788,
"learning_rate": 6.713420275294467e-06,
"loss": 0.6024,
"step": 372
},
{
"epoch": 1.8124049863180298,
"grad_norm": 0.36843356490135193,
"learning_rate": 6.693490148684654e-06,
"loss": 0.572,
"step": 373
},
{
"epoch": 1.817269686834904,
"grad_norm": 0.37639811635017395,
"learning_rate": 6.673529582025398e-06,
"loss": 0.569,
"step": 374
},
{
"epoch": 1.8221343873517788,
"grad_norm": 0.34654489159584045,
"learning_rate": 6.653538934102743e-06,
"loss": 0.6099,
"step": 375
},
{
"epoch": 1.826999087868653,
"grad_norm": 0.3429660499095917,
"learning_rate": 6.633518564243442e-06,
"loss": 0.5892,
"step": 376
},
{
"epoch": 1.8318637883855275,
"grad_norm": 0.3472631275653839,
"learning_rate": 6.6134688323084884e-06,
"loss": 0.5861,
"step": 377
},
{
"epoch": 1.836728488902402,
"grad_norm": 0.40335071086883545,
"learning_rate": 6.593390098686653e-06,
"loss": 0.6067,
"step": 378
},
{
"epoch": 1.8415931894192763,
"grad_norm": 0.37849733233451843,
"learning_rate": 6.573282724288001e-06,
"loss": 0.5956,
"step": 379
},
{
"epoch": 1.8464578899361508,
"grad_norm": 0.360460102558136,
"learning_rate": 6.553147070537413e-06,
"loss": 0.6015,
"step": 380
},
{
"epoch": 1.8513225904530253,
"grad_norm": 0.3695388734340668,
"learning_rate": 6.532983499368078e-06,
"loss": 0.5781,
"step": 381
},
{
"epoch": 1.8561872909698995,
"grad_norm": 0.3779650926589966,
"learning_rate": 6.512792373215e-06,
"loss": 0.5692,
"step": 382
},
{
"epoch": 1.8610519914867742,
"grad_norm": 0.33054274320602417,
"learning_rate": 6.492574055008474e-06,
"loss": 0.5744,
"step": 383
},
{
"epoch": 1.8659166920036485,
"grad_norm": 0.36378973722457886,
"learning_rate": 6.472328908167562e-06,
"loss": 0.586,
"step": 384
},
{
"epoch": 1.870781392520523,
"grad_norm": 0.3403121531009674,
"learning_rate": 6.452057296593568e-06,
"loss": 0.5835,
"step": 385
},
{
"epoch": 1.8756460930373975,
"grad_norm": 0.3431931138038635,
"learning_rate": 6.431759584663492e-06,
"loss": 0.5817,
"step": 386
},
{
"epoch": 1.8805107935542718,
"grad_norm": 0.33876556158065796,
"learning_rate": 6.411436137223479e-06,
"loss": 0.5525,
"step": 387
},
{
"epoch": 1.8853754940711462,
"grad_norm": 0.363239586353302,
"learning_rate": 6.391087319582264e-06,
"loss": 0.5631,
"step": 388
},
{
"epoch": 1.8902401945880207,
"grad_norm": 0.36009472608566284,
"learning_rate": 6.370713497504607e-06,
"loss": 0.5779,
"step": 389
},
{
"epoch": 1.895104895104895,
"grad_norm": 0.357757568359375,
"learning_rate": 6.350315037204714e-06,
"loss": 0.5777,
"step": 390
},
{
"epoch": 1.8999695956217695,
"grad_norm": 0.36344608664512634,
"learning_rate": 6.329892305339659e-06,
"loss": 0.5849,
"step": 391
},
{
"epoch": 1.904834296138644,
"grad_norm": 0.34636056423187256,
"learning_rate": 6.309445669002787e-06,
"loss": 0.5665,
"step": 392
},
{
"epoch": 1.9096989966555182,
"grad_norm": 0.36188262701034546,
"learning_rate": 6.288975495717124e-06,
"loss": 0.5828,
"step": 393
},
{
"epoch": 1.914563697172393,
"grad_norm": 0.34284311532974243,
"learning_rate": 6.268482153428763e-06,
"loss": 0.5872,
"step": 394
},
{
"epoch": 1.9194283976892672,
"grad_norm": 0.37720975279808044,
"learning_rate": 6.247966010500258e-06,
"loss": 0.6029,
"step": 395
},
{
"epoch": 1.9242930982061417,
"grad_norm": 0.35886484384536743,
"learning_rate": 6.227427435703997e-06,
"loss": 0.5892,
"step": 396
},
{
"epoch": 1.9291577987230162,
"grad_norm": 0.3546023666858673,
"learning_rate": 6.206866798215571e-06,
"loss": 0.5784,
"step": 397
},
{
"epoch": 1.9340224992398904,
"grad_norm": 0.3826463520526886,
"learning_rate": 6.186284467607149e-06,
"loss": 0.5936,
"step": 398
},
{
"epoch": 1.938887199756765,
"grad_norm": 0.3731802701950073,
"learning_rate": 6.165680813840822e-06,
"loss": 0.5813,
"step": 399
},
{
"epoch": 1.9437519002736394,
"grad_norm": 0.38415512442588806,
"learning_rate": 6.1450562072619635e-06,
"loss": 0.6,
"step": 400
},
{
"epoch": 1.9486166007905137,
"grad_norm": 0.33490461111068726,
"learning_rate": 6.124411018592568e-06,
"loss": 0.5872,
"step": 401
},
{
"epoch": 1.9534813013073884,
"grad_norm": 0.35179373621940613,
"learning_rate": 6.103745618924587e-06,
"loss": 0.5699,
"step": 402
},
{
"epoch": 1.9583460018242627,
"grad_norm": 0.356187105178833,
"learning_rate": 6.0830603797132574e-06,
"loss": 0.5804,
"step": 403
},
{
"epoch": 1.9632107023411371,
"grad_norm": 0.3447719216346741,
"learning_rate": 6.0623556727704306e-06,
"loss": 0.5828,
"step": 404
},
{
"epoch": 1.9680754028580116,
"grad_norm": 0.3698787987232208,
"learning_rate": 6.041631870257882e-06,
"loss": 0.5897,
"step": 405
},
{
"epoch": 1.972940103374886,
"grad_norm": 0.3562972843647003,
"learning_rate": 6.020889344680627e-06,
"loss": 0.582,
"step": 406
},
{
"epoch": 1.9778048038917604,
"grad_norm": 0.36807456612586975,
"learning_rate": 6.000128468880223e-06,
"loss": 0.5953,
"step": 407
},
{
"epoch": 1.9826695044086349,
"grad_norm": 0.38561809062957764,
"learning_rate": 5.979349616028067e-06,
"loss": 0.58,
"step": 408
},
{
"epoch": 1.9875342049255091,
"grad_norm": 0.3650089204311371,
"learning_rate": 5.958553159618693e-06,
"loss": 0.5925,
"step": 409
},
{
"epoch": 1.9923989054423838,
"grad_norm": 0.35033372044563293,
"learning_rate": 5.937739473463047e-06,
"loss": 0.6011,
"step": 410
},
{
"epoch": 1.9972636059592581,
"grad_norm": 0.3724525570869446,
"learning_rate": 5.916908931681781e-06,
"loss": 0.5716,
"step": 411
},
{
"epoch": 2.0,
"grad_norm": 0.4660073220729828,
"learning_rate": 5.896061908698521e-06,
"loss": 0.5779,
"step": 412
},
{
"epoch": 2.0048647005168743,
"grad_norm": 0.4630298912525177,
"learning_rate": 5.8751987792331365e-06,
"loss": 0.546,
"step": 413
},
{
"epoch": 2.009729401033749,
"grad_norm": 0.4151447117328644,
"learning_rate": 5.854319918295012e-06,
"loss": 0.5555,
"step": 414
},
{
"epoch": 2.0145941015506232,
"grad_norm": 0.4100342094898224,
"learning_rate": 5.833425701176294e-06,
"loss": 0.5728,
"step": 415
},
{
"epoch": 2.0194588020674975,
"grad_norm": 0.4262101352214813,
"learning_rate": 5.812516503445158e-06,
"loss": 0.5753,
"step": 416
},
{
"epoch": 2.024323502584372,
"grad_norm": 0.436073899269104,
"learning_rate": 5.79159270093905e-06,
"loss": 0.5743,
"step": 417
},
{
"epoch": 2.0291882031012465,
"grad_norm": 0.34796959161758423,
"learning_rate": 5.770654669757935e-06,
"loss": 0.5442,
"step": 418
},
{
"epoch": 2.034052903618121,
"grad_norm": 0.38504719734191895,
"learning_rate": 5.749702786257529e-06,
"loss": 0.5322,
"step": 419
},
{
"epoch": 2.0389176041349955,
"grad_norm": 0.3461185693740845,
"learning_rate": 5.7287374270425475e-06,
"loss": 0.5325,
"step": 420
},
{
"epoch": 2.0437823046518697,
"grad_norm": 0.35231730341911316,
"learning_rate": 5.707758968959923e-06,
"loss": 0.5408,
"step": 421
},
{
"epoch": 2.0486470051687444,
"grad_norm": 0.3527251183986664,
"learning_rate": 5.686767789092041e-06,
"loss": 0.5553,
"step": 422
},
{
"epoch": 2.0535117056856187,
"grad_norm": 0.37650537490844727,
"learning_rate": 5.6657642647499545e-06,
"loss": 0.5758,
"step": 423
},
{
"epoch": 2.058376406202493,
"grad_norm": 0.34170427918434143,
"learning_rate": 5.644748773466606e-06,
"loss": 0.5278,
"step": 424
},
{
"epoch": 2.0632411067193677,
"grad_norm": 0.3250020146369934,
"learning_rate": 5.62372169299004e-06,
"loss": 0.5496,
"step": 425
},
{
"epoch": 2.068105807236242,
"grad_norm": 0.36646807193756104,
"learning_rate": 5.6026834012766155e-06,
"loss": 0.5431,
"step": 426
},
{
"epoch": 2.0729705077531166,
"grad_norm": 0.38466954231262207,
"learning_rate": 5.581634276484211e-06,
"loss": 0.5646,
"step": 427
},
{
"epoch": 2.077835208269991,
"grad_norm": 0.3858664929866791,
"learning_rate": 5.560574696965425e-06,
"loss": 0.5789,
"step": 428
},
{
"epoch": 2.082699908786865,
"grad_norm": 0.36406323313713074,
"learning_rate": 5.539505041260779e-06,
"loss": 0.5464,
"step": 429
},
{
"epoch": 2.08756460930374,
"grad_norm": 0.3602047264575958,
"learning_rate": 5.518425688091906e-06,
"loss": 0.5615,
"step": 430
},
{
"epoch": 2.092429309820614,
"grad_norm": 0.3504374623298645,
"learning_rate": 5.497337016354757e-06,
"loss": 0.5608,
"step": 431
},
{
"epoch": 2.0972940103374884,
"grad_norm": 0.3517240285873413,
"learning_rate": 5.476239405112775e-06,
"loss": 0.5437,
"step": 432
},
{
"epoch": 2.102158710854363,
"grad_norm": 0.3447907865047455,
"learning_rate": 5.45513323359009e-06,
"loss": 0.5514,
"step": 433
},
{
"epoch": 2.1070234113712374,
"grad_norm": 0.34815070033073425,
"learning_rate": 5.434018881164702e-06,
"loss": 0.5374,
"step": 434
},
{
"epoch": 2.111888111888112,
"grad_norm": 0.37003400921821594,
"learning_rate": 5.412896727361663e-06,
"loss": 0.5301,
"step": 435
},
{
"epoch": 2.1167528124049864,
"grad_norm": 0.32090049982070923,
"learning_rate": 5.391767151846247e-06,
"loss": 0.5157,
"step": 436
},
{
"epoch": 2.1216175129218606,
"grad_norm": 0.3075619339942932,
"learning_rate": 5.370630534417133e-06,
"loss": 0.5393,
"step": 437
},
{
"epoch": 2.1264822134387353,
"grad_norm": 0.35239556431770325,
"learning_rate": 5.349487254999579e-06,
"loss": 0.5588,
"step": 438
},
{
"epoch": 2.1313469139556096,
"grad_norm": 0.37438079714775085,
"learning_rate": 5.328337693638591e-06,
"loss": 0.5507,
"step": 439
},
{
"epoch": 2.136211614472484,
"grad_norm": 0.3725924789905548,
"learning_rate": 5.307182230492089e-06,
"loss": 0.5674,
"step": 440
},
{
"epoch": 2.1410763149893586,
"grad_norm": 0.34325110912323,
"learning_rate": 5.286021245824075e-06,
"loss": 0.5463,
"step": 441
},
{
"epoch": 2.145941015506233,
"grad_norm": 0.3551580011844635,
"learning_rate": 5.264855119997803e-06,
"loss": 0.555,
"step": 442
},
{
"epoch": 2.1508057160231075,
"grad_norm": 0.3694038987159729,
"learning_rate": 5.243684233468933e-06,
"loss": 0.5526,
"step": 443
},
{
"epoch": 2.155670416539982,
"grad_norm": 0.3545861542224884,
"learning_rate": 5.222508966778702e-06,
"loss": 0.5688,
"step": 444
},
{
"epoch": 2.160535117056856,
"grad_norm": 0.39304155111312866,
"learning_rate": 5.201329700547077e-06,
"loss": 0.5612,
"step": 445
},
{
"epoch": 2.165399817573731,
"grad_norm": 0.3505667448043823,
"learning_rate": 5.180146815465915e-06,
"loss": 0.5468,
"step": 446
},
{
"epoch": 2.170264518090605,
"grad_norm": 0.3596620559692383,
"learning_rate": 5.158960692292122e-06,
"loss": 0.5471,
"step": 447
},
{
"epoch": 2.1751292186074793,
"grad_norm": 0.33798959851264954,
"learning_rate": 5.137771711840811e-06,
"loss": 0.5528,
"step": 448
},
{
"epoch": 2.179993919124354,
"grad_norm": 0.37839508056640625,
"learning_rate": 5.116580254978447e-06,
"loss": 0.5476,
"step": 449
},
{
"epoch": 2.1848586196412283,
"grad_norm": 0.35490870475769043,
"learning_rate": 5.095386702616012e-06,
"loss": 0.5551,
"step": 450
},
{
"epoch": 2.1897233201581026,
"grad_norm": 0.3636198937892914,
"learning_rate": 5.074191435702155e-06,
"loss": 0.5547,
"step": 451
},
{
"epoch": 2.1945880206749773,
"grad_norm": 0.36096078157424927,
"learning_rate": 5.05299483521634e-06,
"loss": 0.5694,
"step": 452
},
{
"epoch": 2.1994527211918515,
"grad_norm": 0.3548087179660797,
"learning_rate": 5.031797282162007e-06,
"loss": 0.5591,
"step": 453
},
{
"epoch": 2.2043174217087262,
"grad_norm": 0.38166943192481995,
"learning_rate": 5.010599157559713e-06,
"loss": 0.5406,
"step": 454
},
{
"epoch": 2.2091821222256005,
"grad_norm": 0.3304060399532318,
"learning_rate": 4.98940084244029e-06,
"loss": 0.5525,
"step": 455
},
{
"epoch": 2.2140468227424748,
"grad_norm": 0.35069623589515686,
"learning_rate": 4.968202717837996e-06,
"loss": 0.5418,
"step": 456
},
{
"epoch": 2.2189115232593495,
"grad_norm": 0.3588007092475891,
"learning_rate": 4.947005164783661e-06,
"loss": 0.5337,
"step": 457
},
{
"epoch": 2.2237762237762237,
"grad_norm": 0.36242586374282837,
"learning_rate": 4.925808564297847e-06,
"loss": 0.5535,
"step": 458
},
{
"epoch": 2.228640924293098,
"grad_norm": 0.33744388818740845,
"learning_rate": 4.9046132973839895e-06,
"loss": 0.546,
"step": 459
},
{
"epoch": 2.2335056248099727,
"grad_norm": 0.3615228831768036,
"learning_rate": 4.883419745021554e-06,
"loss": 0.5483,
"step": 460
},
{
"epoch": 2.238370325326847,
"grad_norm": 0.3409022390842438,
"learning_rate": 4.862228288159191e-06,
"loss": 0.5428,
"step": 461
},
{
"epoch": 2.2432350258437217,
"grad_norm": 0.35754162073135376,
"learning_rate": 4.841039307707878e-06,
"loss": 0.5471,
"step": 462
},
{
"epoch": 2.248099726360596,
"grad_norm": 0.3479742109775543,
"learning_rate": 4.819853184534085e-06,
"loss": 0.5432,
"step": 463
},
{
"epoch": 2.2529644268774702,
"grad_norm": 0.3466337323188782,
"learning_rate": 4.798670299452926e-06,
"loss": 0.5302,
"step": 464
},
{
"epoch": 2.257829127394345,
"grad_norm": 0.37863853573799133,
"learning_rate": 4.7774910332213005e-06,
"loss": 0.53,
"step": 465
},
{
"epoch": 2.262693827911219,
"grad_norm": 0.4052421748638153,
"learning_rate": 4.756315766531069e-06,
"loss": 0.5983,
"step": 466
},
{
"epoch": 2.2675585284280935,
"grad_norm": 0.3390233814716339,
"learning_rate": 4.735144880002199e-06,
"loss": 0.547,
"step": 467
},
{
"epoch": 2.272423228944968,
"grad_norm": 0.3182123601436615,
"learning_rate": 4.713978754175926e-06,
"loss": 0.5284,
"step": 468
},
{
"epoch": 2.2772879294618424,
"grad_norm": 0.3559403419494629,
"learning_rate": 4.692817769507912e-06,
"loss": 0.564,
"step": 469
},
{
"epoch": 2.282152629978717,
"grad_norm": 0.3409214913845062,
"learning_rate": 4.671662306361409e-06,
"loss": 0.5582,
"step": 470
},
{
"epoch": 2.2870173304955914,
"grad_norm": 0.3738288879394531,
"learning_rate": 4.6505127450004216e-06,
"loss": 0.5723,
"step": 471
},
{
"epoch": 2.2918820310124657,
"grad_norm": 0.34341713786125183,
"learning_rate": 4.62936946558287e-06,
"loss": 0.5477,
"step": 472
},
{
"epoch": 2.2967467315293404,
"grad_norm": 0.34464555978775024,
"learning_rate": 4.608232848153757e-06,
"loss": 0.5692,
"step": 473
},
{
"epoch": 2.3016114320462147,
"grad_norm": 0.3629922568798065,
"learning_rate": 4.587103272638339e-06,
"loss": 0.5479,
"step": 474
},
{
"epoch": 2.306476132563089,
"grad_norm": 0.32645168900489807,
"learning_rate": 4.565981118835299e-06,
"loss": 0.5783,
"step": 475
},
{
"epoch": 2.3113408330799636,
"grad_norm": 0.3384042978286743,
"learning_rate": 4.5448667664099125e-06,
"loss": 0.5463,
"step": 476
},
{
"epoch": 2.316205533596838,
"grad_norm": 0.36268654465675354,
"learning_rate": 4.523760594887228e-06,
"loss": 0.5696,
"step": 477
},
{
"epoch": 2.321070234113712,
"grad_norm": 0.33541953563690186,
"learning_rate": 4.5026629836452445e-06,
"loss": 0.5544,
"step": 478
},
{
"epoch": 2.325934934630587,
"grad_norm": 0.3592680096626282,
"learning_rate": 4.481574311908096e-06,
"loss": 0.5411,
"step": 479
},
{
"epoch": 2.330799635147461,
"grad_norm": 0.3406287431716919,
"learning_rate": 4.460494958739223e-06,
"loss": 0.5415,
"step": 480
},
{
"epoch": 2.335664335664336,
"grad_norm": 0.34424182772636414,
"learning_rate": 4.439425303034576e-06,
"loss": 0.5498,
"step": 481
},
{
"epoch": 2.34052903618121,
"grad_norm": 0.38854897022247314,
"learning_rate": 4.418365723515791e-06,
"loss": 0.5604,
"step": 482
},
{
"epoch": 2.3453937366980844,
"grad_norm": 0.3248487412929535,
"learning_rate": 4.397316598723385e-06,
"loss": 0.5667,
"step": 483
},
{
"epoch": 2.350258437214959,
"grad_norm": 0.38373878598213196,
"learning_rate": 4.376278307009962e-06,
"loss": 0.5714,
"step": 484
},
{
"epoch": 2.3551231377318333,
"grad_norm": 0.35804232954978943,
"learning_rate": 4.355251226533396e-06,
"loss": 0.5583,
"step": 485
},
{
"epoch": 2.359987838248708,
"grad_norm": 0.3704029619693756,
"learning_rate": 4.334235735250047e-06,
"loss": 0.5389,
"step": 486
},
{
"epoch": 2.3648525387655823,
"grad_norm": 0.3661844730377197,
"learning_rate": 4.313232210907959e-06,
"loss": 0.5484,
"step": 487
},
{
"epoch": 2.3697172392824566,
"grad_norm": 0.3929769992828369,
"learning_rate": 4.292241031040077e-06,
"loss": 0.5455,
"step": 488
},
{
"epoch": 2.374581939799331,
"grad_norm": 0.36609140038490295,
"learning_rate": 4.271262572957453e-06,
"loss": 0.5462,
"step": 489
},
{
"epoch": 2.3794466403162056,
"grad_norm": 0.36633870005607605,
"learning_rate": 4.250297213742473e-06,
"loss": 0.5531,
"step": 490
},
{
"epoch": 2.38431134083308,
"grad_norm": 0.3468073308467865,
"learning_rate": 4.229345330242067e-06,
"loss": 0.5392,
"step": 491
},
{
"epoch": 2.3891760413499545,
"grad_norm": 0.36657583713531494,
"learning_rate": 4.2084072990609505e-06,
"loss": 0.5818,
"step": 492
},
{
"epoch": 2.394040741866829,
"grad_norm": 0.35404083132743835,
"learning_rate": 4.187483496554844e-06,
"loss": 0.5466,
"step": 493
},
{
"epoch": 2.398905442383703,
"grad_norm": 0.3507364094257355,
"learning_rate": 4.166574298823707e-06,
"loss": 0.556,
"step": 494
},
{
"epoch": 2.4037701429005778,
"grad_norm": 0.3688286244869232,
"learning_rate": 4.145680081704989e-06,
"loss": 0.5722,
"step": 495
},
{
"epoch": 2.408634843417452,
"grad_norm": 0.4607541561126709,
"learning_rate": 4.1248012207668635e-06,
"loss": 0.5614,
"step": 496
},
{
"epoch": 2.4134995439343268,
"grad_norm": 0.34164538979530334,
"learning_rate": 4.103938091301479e-06,
"loss": 0.5411,
"step": 497
},
{
"epoch": 2.418364244451201,
"grad_norm": 0.37179362773895264,
"learning_rate": 4.08309106831822e-06,
"loss": 0.5697,
"step": 498
},
{
"epoch": 2.4232289449680753,
"grad_norm": 0.3777942359447479,
"learning_rate": 4.062260526536955e-06,
"loss": 0.5479,
"step": 499
},
{
"epoch": 2.42809364548495,
"grad_norm": 0.3554550111293793,
"learning_rate": 4.041446840381309e-06,
"loss": 0.5559,
"step": 500
},
{
"epoch": 2.4329583460018243,
"grad_norm": 0.3182331919670105,
"learning_rate": 4.0206503839719335e-06,
"loss": 0.5224,
"step": 501
},
{
"epoch": 2.4378230465186985,
"grad_norm": 0.3521728515625,
"learning_rate": 3.999871531119779e-06,
"loss": 0.5684,
"step": 502
},
{
"epoch": 2.4426877470355732,
"grad_norm": 0.35995247960090637,
"learning_rate": 3.9791106553193746e-06,
"loss": 0.532,
"step": 503
},
{
"epoch": 2.4475524475524475,
"grad_norm": 0.40925824642181396,
"learning_rate": 3.9583681297421194e-06,
"loss": 0.5771,
"step": 504
},
{
"epoch": 2.4524171480693218,
"grad_norm": 0.37918710708618164,
"learning_rate": 3.937644327229572e-06,
"loss": 0.5447,
"step": 505
},
{
"epoch": 2.4572818485861965,
"grad_norm": 0.33770713210105896,
"learning_rate": 3.916939620286743e-06,
"loss": 0.573,
"step": 506
},
{
"epoch": 2.4621465491030707,
"grad_norm": 0.334839403629303,
"learning_rate": 3.896254381075416e-06,
"loss": 0.5683,
"step": 507
},
{
"epoch": 2.4670112496199454,
"grad_norm": 0.3860621154308319,
"learning_rate": 3.875588981407433e-06,
"loss": 0.5359,
"step": 508
},
{
"epoch": 2.4718759501368197,
"grad_norm": 0.3286064565181732,
"learning_rate": 3.854943792738037e-06,
"loss": 0.5502,
"step": 509
},
{
"epoch": 2.476740650653694,
"grad_norm": 0.38320592045783997,
"learning_rate": 3.834319186159179e-06,
"loss": 0.5624,
"step": 510
},
{
"epoch": 2.4816053511705687,
"grad_norm": 0.3666168749332428,
"learning_rate": 3.8137155323928526e-06,
"loss": 0.5628,
"step": 511
},
{
"epoch": 2.486470051687443,
"grad_norm": 0.35792431235313416,
"learning_rate": 3.7931332017844302e-06,
"loss": 0.5677,
"step": 512
},
{
"epoch": 2.4913347522043177,
"grad_norm": 0.3512524962425232,
"learning_rate": 3.7725725642960047e-06,
"loss": 0.5564,
"step": 513
},
{
"epoch": 2.496199452721192,
"grad_norm": 0.3470909595489502,
"learning_rate": 3.752033989499742e-06,
"loss": 0.5456,
"step": 514
},
{
"epoch": 2.501064153238066,
"grad_norm": 0.350885808467865,
"learning_rate": 3.7315178465712364e-06,
"loss": 0.5485,
"step": 515
},
{
"epoch": 2.5059288537549405,
"grad_norm": 0.34219297766685486,
"learning_rate": 3.7110245042828786e-06,
"loss": 0.5693,
"step": 516
},
{
"epoch": 2.510793554271815,
"grad_norm": 0.38096287846565247,
"learning_rate": 3.690554330997215e-06,
"loss": 0.5416,
"step": 517
},
{
"epoch": 2.5156582547886894,
"grad_norm": 0.3562045693397522,
"learning_rate": 3.670107694660343e-06,
"loss": 0.5566,
"step": 518
},
{
"epoch": 2.520522955305564,
"grad_norm": 0.3100724220275879,
"learning_rate": 3.6496849627952875e-06,
"loss": 0.5407,
"step": 519
},
{
"epoch": 2.5253876558224384,
"grad_norm": 0.360066294670105,
"learning_rate": 3.6292865024953945e-06,
"loss": 0.5829,
"step": 520
},
{
"epoch": 2.5302523563393127,
"grad_norm": 0.34163594245910645,
"learning_rate": 3.6089126804177373e-06,
"loss": 0.5498,
"step": 521
},
{
"epoch": 2.5351170568561874,
"grad_norm": 0.3224615752696991,
"learning_rate": 3.5885638627765228e-06,
"loss": 0.5584,
"step": 522
},
{
"epoch": 2.5399817573730616,
"grad_norm": 0.31611204147338867,
"learning_rate": 3.568240415336509e-06,
"loss": 0.5496,
"step": 523
},
{
"epoch": 2.5448464578899364,
"grad_norm": 0.35780468583106995,
"learning_rate": 3.547942703406433e-06,
"loss": 0.5533,
"step": 524
},
{
"epoch": 2.5497111584068106,
"grad_norm": 0.3295901119709015,
"learning_rate": 3.52767109183244e-06,
"loss": 0.5207,
"step": 525
},
{
"epoch": 2.554575858923685,
"grad_norm": 0.6567420959472656,
"learning_rate": 3.507425944991529e-06,
"loss": 0.5502,
"step": 526
},
{
"epoch": 2.5594405594405596,
"grad_norm": 0.33516809344291687,
"learning_rate": 3.4872076267850015e-06,
"loss": 0.5243,
"step": 527
},
{
"epoch": 2.564305259957434,
"grad_norm": 0.34328529238700867,
"learning_rate": 3.4670165006319236e-06,
"loss": 0.5442,
"step": 528
},
{
"epoch": 2.5691699604743086,
"grad_norm": 0.3590414524078369,
"learning_rate": 3.4468529294625895e-06,
"loss": 0.5503,
"step": 529
},
{
"epoch": 2.574034660991183,
"grad_norm": 0.3209679424762726,
"learning_rate": 3.4267172757120005e-06,
"loss": 0.565,
"step": 530
},
{
"epoch": 2.578899361508057,
"grad_norm": 0.34146201610565186,
"learning_rate": 3.406609901313349e-06,
"loss": 0.5542,
"step": 531
},
{
"epoch": 2.5837640620249314,
"grad_norm": 0.39411503076553345,
"learning_rate": 3.386531167691512e-06,
"loss": 0.5551,
"step": 532
},
{
"epoch": 2.588628762541806,
"grad_norm": 0.3769046664237976,
"learning_rate": 3.36648143575656e-06,
"loss": 0.5556,
"step": 533
},
{
"epoch": 2.5934934630586803,
"grad_norm": 0.3681861460208893,
"learning_rate": 3.3464610658972584e-06,
"loss": 0.5491,
"step": 534
},
{
"epoch": 2.598358163575555,
"grad_norm": 0.3568417429924011,
"learning_rate": 3.326470417974604e-06,
"loss": 0.5363,
"step": 535
},
{
"epoch": 2.6032228640924293,
"grad_norm": 0.3567221760749817,
"learning_rate": 3.3065098513153473e-06,
"loss": 0.5241,
"step": 536
},
{
"epoch": 2.6080875646093036,
"grad_norm": 0.37171709537506104,
"learning_rate": 3.2865797247055354e-06,
"loss": 0.5781,
"step": 537
},
{
"epoch": 2.6129522651261783,
"grad_norm": 0.3496547043323517,
"learning_rate": 3.266680396384061e-06,
"loss": 0.5441,
"step": 538
},
{
"epoch": 2.6178169656430526,
"grad_norm": 0.3279050290584564,
"learning_rate": 3.2468122240362287e-06,
"loss": 0.5576,
"step": 539
},
{
"epoch": 2.6226816661599273,
"grad_norm": 0.33857056498527527,
"learning_rate": 3.226975564787322e-06,
"loss": 0.5564,
"step": 540
},
{
"epoch": 2.6275463666768015,
"grad_norm": 0.3338910937309265,
"learning_rate": 3.2071707751961838e-06,
"loss": 0.5534,
"step": 541
},
{
"epoch": 2.632411067193676,
"grad_norm": 0.35142219066619873,
"learning_rate": 3.187398211248811e-06,
"loss": 0.5426,
"step": 542
},
{
"epoch": 2.63727576771055,
"grad_norm": 0.3173449635505676,
"learning_rate": 3.1676582283519454e-06,
"loss": 0.5452,
"step": 543
},
{
"epoch": 2.6421404682274248,
"grad_norm": 0.37744569778442383,
"learning_rate": 3.1479511813267006e-06,
"loss": 0.5448,
"step": 544
},
{
"epoch": 2.647005168744299,
"grad_norm": 0.3497507870197296,
"learning_rate": 3.1282774244021717e-06,
"loss": 0.5357,
"step": 545
},
{
"epoch": 2.6518698692611737,
"grad_norm": 0.33956778049468994,
"learning_rate": 3.1086373112090762e-06,
"loss": 0.5683,
"step": 546
},
{
"epoch": 2.656734569778048,
"grad_norm": 0.30914613604545593,
"learning_rate": 3.089031194773392e-06,
"loss": 0.5614,
"step": 547
},
{
"epoch": 2.6615992702949223,
"grad_norm": 0.34617817401885986,
"learning_rate": 3.069459427510014e-06,
"loss": 0.566,
"step": 548
},
{
"epoch": 2.666463970811797,
"grad_norm": 0.3441532850265503,
"learning_rate": 3.049922361216422e-06,
"loss": 0.5493,
"step": 549
},
{
"epoch": 2.6713286713286712,
"grad_norm": 0.34598854184150696,
"learning_rate": 3.0304203470663507e-06,
"loss": 0.5466,
"step": 550
},
{
"epoch": 2.676193371845546,
"grad_norm": 0.3242530822753906,
"learning_rate": 3.0109537356034856e-06,
"loss": 0.5501,
"step": 551
},
{
"epoch": 2.68105807236242,
"grad_norm": 0.3374408781528473,
"learning_rate": 2.991522876735154e-06,
"loss": 0.5386,
"step": 552
},
{
"epoch": 2.6859227728792945,
"grad_norm": 0.33738142251968384,
"learning_rate": 2.9721281197260427e-06,
"loss": 0.5648,
"step": 553
},
{
"epoch": 2.690787473396169,
"grad_norm": 0.3745483458042145,
"learning_rate": 2.9527698131919156e-06,
"loss": 0.5637,
"step": 554
},
{
"epoch": 2.6956521739130435,
"grad_norm": 0.36665889620780945,
"learning_rate": 2.9334483050933506e-06,
"loss": 0.5443,
"step": 555
},
{
"epoch": 2.700516874429918,
"grad_norm": 0.3387942314147949,
"learning_rate": 2.91416394272948e-06,
"loss": 0.558,
"step": 556
},
{
"epoch": 2.7053815749467924,
"grad_norm": 0.3336210548877716,
"learning_rate": 2.894917072731753e-06,
"loss": 0.5441,
"step": 557
},
{
"epoch": 2.7102462754636667,
"grad_norm": 0.3553939759731293,
"learning_rate": 2.8757080410577042e-06,
"loss": 0.5706,
"step": 558
},
{
"epoch": 2.715110975980541,
"grad_norm": 0.328545480966568,
"learning_rate": 2.8565371929847286e-06,
"loss": 0.5297,
"step": 559
},
{
"epoch": 2.7199756764974157,
"grad_norm": 0.35486361384391785,
"learning_rate": 2.83740487310389e-06,
"loss": 0.5432,
"step": 560
},
{
"epoch": 2.72484037701429,
"grad_norm": 0.34170064330101013,
"learning_rate": 2.81831142531371e-06,
"loss": 0.55,
"step": 561
},
{
"epoch": 2.7297050775311646,
"grad_norm": 0.3448440432548523,
"learning_rate": 2.7992571928139984e-06,
"loss": 0.5778,
"step": 562
},
{
"epoch": 2.734569778048039,
"grad_norm": 0.347851037979126,
"learning_rate": 2.780242518099675e-06,
"loss": 0.5749,
"step": 563
},
{
"epoch": 2.739434478564913,
"grad_norm": 0.3358885645866394,
"learning_rate": 2.761267742954629e-06,
"loss": 0.5406,
"step": 564
},
{
"epoch": 2.744299179081788,
"grad_norm": 0.3412470817565918,
"learning_rate": 2.7423332084455543e-06,
"loss": 0.5644,
"step": 565
},
{
"epoch": 2.749163879598662,
"grad_norm": 0.35414665937423706,
"learning_rate": 2.723439254915834e-06,
"loss": 0.571,
"step": 566
},
{
"epoch": 2.754028580115537,
"grad_norm": 0.33787044882774353,
"learning_rate": 2.704586221979422e-06,
"loss": 0.5707,
"step": 567
},
{
"epoch": 2.758893280632411,
"grad_norm": 0.3428189158439636,
"learning_rate": 2.6857744485147286e-06,
"loss": 0.5348,
"step": 568
},
{
"epoch": 2.7637579811492854,
"grad_norm": 0.33465680480003357,
"learning_rate": 2.667004272658541e-06,
"loss": 0.5413,
"step": 569
},
{
"epoch": 2.7686226816661597,
"grad_norm": 0.3357372581958771,
"learning_rate": 2.6482760317999338e-06,
"loss": 0.5649,
"step": 570
},
{
"epoch": 2.7734873821830344,
"grad_norm": 0.3477322459220886,
"learning_rate": 2.629590062574221e-06,
"loss": 0.5629,
"step": 571
},
{
"epoch": 2.7783520826999086,
"grad_norm": 0.3558913469314575,
"learning_rate": 2.610946700856885e-06,
"loss": 0.5673,
"step": 572
},
{
"epoch": 2.7832167832167833,
"grad_norm": 0.30213209986686707,
"learning_rate": 2.592346281757552e-06,
"loss": 0.5606,
"step": 573
},
{
"epoch": 2.7880814837336576,
"grad_norm": 0.3280998766422272,
"learning_rate": 2.5737891396139713e-06,
"loss": 0.5407,
"step": 574
},
{
"epoch": 2.792946184250532,
"grad_norm": 0.32864999771118164,
"learning_rate": 2.5552756079859904e-06,
"loss": 0.545,
"step": 575
},
{
"epoch": 2.7978108847674066,
"grad_norm": 0.3435007631778717,
"learning_rate": 2.5368060196495785e-06,
"loss": 0.5334,
"step": 576
},
{
"epoch": 2.802675585284281,
"grad_norm": 0.35111281275749207,
"learning_rate": 2.5183807065908296e-06,
"loss": 0.5587,
"step": 577
},
{
"epoch": 2.8075402858011556,
"grad_norm": 0.3694814443588257,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.5478,
"step": 578
},
{
"epoch": 2.81240498631803,
"grad_norm": 0.32266682386398315,
"learning_rate": 2.4816642302655634e-06,
"loss": 0.5337,
"step": 579
},
{
"epoch": 2.817269686834904,
"grad_norm": 0.3546319007873535,
"learning_rate": 2.4633737269682546e-06,
"loss": 0.5694,
"step": 580
},
{
"epoch": 2.822134387351779,
"grad_norm": 0.3136415481567383,
"learning_rate": 2.445128818875166e-06,
"loss": 0.551,
"step": 581
},
{
"epoch": 2.826999087868653,
"grad_norm": 0.3577171564102173,
"learning_rate": 2.4269298339338205e-06,
"loss": 0.5558,
"step": 582
},
{
"epoch": 2.8318637883855278,
"grad_norm": 0.3270150423049927,
"learning_rate": 2.408777099266291e-06,
"loss": 0.5357,
"step": 583
},
{
"epoch": 2.836728488902402,
"grad_norm": 0.3563329875469208,
"learning_rate": 2.3906709411633073e-06,
"loss": 0.5487,
"step": 584
},
{
"epoch": 2.8415931894192763,
"grad_norm": 0.347988486289978,
"learning_rate": 2.3726116850783987e-06,
"loss": 0.533,
"step": 585
},
{
"epoch": 2.8464578899361506,
"grad_norm": 0.3296806812286377,
"learning_rate": 2.354599655622049e-06,
"loss": 0.5708,
"step": 586
},
{
"epoch": 2.8513225904530253,
"grad_norm": 0.32138484716415405,
"learning_rate": 2.3366351765558437e-06,
"loss": 0.5431,
"step": 587
},
{
"epoch": 2.8561872909698995,
"grad_norm": 0.33454257249832153,
"learning_rate": 2.318718570786675e-06,
"loss": 0.5216,
"step": 588
},
{
"epoch": 2.8610519914867742,
"grad_norm": 0.34356728196144104,
"learning_rate": 2.3008501603609147e-06,
"loss": 0.541,
"step": 589
},
{
"epoch": 2.8659166920036485,
"grad_norm": 0.3413681089878082,
"learning_rate": 2.283030266458644e-06,
"loss": 0.5065,
"step": 590
},
{
"epoch": 2.870781392520523,
"grad_norm": 0.31537744402885437,
"learning_rate": 2.265259209387867e-06,
"loss": 0.5633,
"step": 591
},
{
"epoch": 2.8756460930373975,
"grad_norm": 0.3271123468875885,
"learning_rate": 2.2475373085787568e-06,
"loss": 0.5442,
"step": 592
},
{
"epoch": 2.8805107935542718,
"grad_norm": 0.3457365930080414,
"learning_rate": 2.229864882577921e-06,
"loss": 0.5706,
"step": 593
},
{
"epoch": 2.8853754940711465,
"grad_norm": 0.3140386939048767,
"learning_rate": 2.2122422490426676e-06,
"loss": 0.5467,
"step": 594
},
{
"epoch": 2.8902401945880207,
"grad_norm": 0.32702910900115967,
"learning_rate": 2.194669724735296e-06,
"loss": 0.5515,
"step": 595
},
{
"epoch": 2.895104895104895,
"grad_norm": 0.3251449465751648,
"learning_rate": 2.1771476255174056e-06,
"loss": 0.5365,
"step": 596
},
{
"epoch": 2.8999695956217693,
"grad_norm": 0.33286523818969727,
"learning_rate": 2.159676266344222e-06,
"loss": 0.5383,
"step": 597
},
{
"epoch": 2.904834296138644,
"grad_norm": 0.33338719606399536,
"learning_rate": 2.1422559612589266e-06,
"loss": 0.5532,
"step": 598
},
{
"epoch": 2.9096989966555182,
"grad_norm": 0.33476975560188293,
"learning_rate": 2.124887023387017e-06,
"loss": 0.5638,
"step": 599
},
{
"epoch": 2.914563697172393,
"grad_norm": 0.3562331199645996,
"learning_rate": 2.1075697649306838e-06,
"loss": 0.5463,
"step": 600
},
{
"epoch": 2.919428397689267,
"grad_norm": 0.3595515191555023,
"learning_rate": 2.0903044971631854e-06,
"loss": 0.5608,
"step": 601
},
{
"epoch": 2.9242930982061415,
"grad_norm": 0.34842994809150696,
"learning_rate": 2.0730915304232692e-06,
"loss": 0.5482,
"step": 602
},
{
"epoch": 2.929157798723016,
"grad_norm": 0.31734150648117065,
"learning_rate": 2.055931174109579e-06,
"loss": 0.5313,
"step": 603
},
{
"epoch": 2.9340224992398904,
"grad_norm": 0.3557087481021881,
"learning_rate": 2.0388237366751005e-06,
"loss": 0.54,
"step": 604
},
{
"epoch": 2.938887199756765,
"grad_norm": 0.3495935797691345,
"learning_rate": 2.02176952562162e-06,
"loss": 0.5434,
"step": 605
},
{
"epoch": 2.9437519002736394,
"grad_norm": 0.3390326499938965,
"learning_rate": 2.004768847494186e-06,
"loss": 0.5492,
"step": 606
},
{
"epoch": 2.9486166007905137,
"grad_norm": 0.33091801404953003,
"learning_rate": 1.987822007875617e-06,
"loss": 0.5631,
"step": 607
},
{
"epoch": 2.9534813013073884,
"grad_norm": 0.3437463045120239,
"learning_rate": 1.9709293113809876e-06,
"loss": 0.5654,
"step": 608
},
{
"epoch": 2.9583460018242627,
"grad_norm": 0.3187878727912903,
"learning_rate": 1.954091061652172e-06,
"loss": 0.5606,
"step": 609
},
{
"epoch": 2.9632107023411374,
"grad_norm": 0.3530491590499878,
"learning_rate": 1.9373075613523728e-06,
"loss": 0.559,
"step": 610
},
{
"epoch": 2.9680754028580116,
"grad_norm": 0.3447261154651642,
"learning_rate": 1.920579112160685e-06,
"loss": 0.5541,
"step": 611
},
{
"epoch": 2.972940103374886,
"grad_norm": 0.315826952457428,
"learning_rate": 1.903906014766681e-06,
"loss": 0.5472,
"step": 612
},
{
"epoch": 2.97780480389176,
"grad_norm": 0.3575029969215393,
"learning_rate": 1.8872885688649879e-06,
"loss": 0.5399,
"step": 613
},
{
"epoch": 2.982669504408635,
"grad_norm": 0.338563472032547,
"learning_rate": 1.8707270731499223e-06,
"loss": 0.5509,
"step": 614
},
{
"epoch": 2.987534204925509,
"grad_norm": 0.31489479541778564,
"learning_rate": 1.854221825310103e-06,
"loss": 0.5498,
"step": 615
},
{
"epoch": 2.992398905442384,
"grad_norm": 0.3175964951515198,
"learning_rate": 1.8377731220231144e-06,
"loss": 0.5532,
"step": 616
},
{
"epoch": 2.997263605959258,
"grad_norm": 0.3167542517185211,
"learning_rate": 1.8213812589501611e-06,
"loss": 0.5598,
"step": 617
},
{
"epoch": 3.0,
"grad_norm": 0.45881038904190063,
"learning_rate": 1.8050465307307602e-06,
"loss": 0.5569,
"step": 618
},
{
"epoch": 3.0048647005168743,
"grad_norm": 0.39067137241363525,
"learning_rate": 1.78876923097745e-06,
"loss": 0.5258,
"step": 619
},
{
"epoch": 3.009729401033749,
"grad_norm": 0.40629592537879944,
"learning_rate": 1.7725496522704998e-06,
"loss": 0.549,
"step": 620
},
{
"epoch": 3.0145941015506232,
"grad_norm": 0.3575362265110016,
"learning_rate": 1.7563880861526656e-06,
"loss": 0.5368,
"step": 621
},
{
"epoch": 3.0194588020674975,
"grad_norm": 0.3887767493724823,
"learning_rate": 1.7402848231239317e-06,
"loss": 0.5526,
"step": 622
},
{
"epoch": 3.024323502584372,
"grad_norm": 0.33983156085014343,
"learning_rate": 1.7242401526363095e-06,
"loss": 0.5423,
"step": 623
},
{
"epoch": 3.0291882031012465,
"grad_norm": 0.3710326552391052,
"learning_rate": 1.70825436308862e-06,
"loss": 0.5402,
"step": 624
},
{
"epoch": 3.034052903618121,
"grad_norm": 0.3933396339416504,
"learning_rate": 1.692327741821312e-06,
"loss": 0.535,
"step": 625
},
{
"epoch": 3.0389176041349955,
"grad_norm": 0.358233779668808,
"learning_rate": 1.676460575111306e-06,
"loss": 0.5327,
"step": 626
},
{
"epoch": 3.0437823046518697,
"grad_norm": 0.3487701714038849,
"learning_rate": 1.6606531481668364e-06,
"loss": 0.5441,
"step": 627
},
{
"epoch": 3.0486470051687444,
"grad_norm": 0.32701754570007324,
"learning_rate": 1.6449057451223354e-06,
"loss": 0.5171,
"step": 628
},
{
"epoch": 3.0535117056856187,
"grad_norm": 0.31877729296684265,
"learning_rate": 1.6292186490333172e-06,
"loss": 0.5332,
"step": 629
},
{
"epoch": 3.058376406202493,
"grad_norm": 0.3283981680870056,
"learning_rate": 1.6135921418712959e-06,
"loss": 0.5429,
"step": 630
},
{
"epoch": 3.0632411067193677,
"grad_norm": 0.3563496172428131,
"learning_rate": 1.5980265045187139e-06,
"loss": 0.5424,
"step": 631
},
{
"epoch": 3.068105807236242,
"grad_norm": 0.3167465031147003,
"learning_rate": 1.5825220167638945e-06,
"loss": 0.5179,
"step": 632
},
{
"epoch": 3.0729705077531166,
"grad_norm": 0.37663573026657104,
"learning_rate": 1.567078957296016e-06,
"loss": 0.552,
"step": 633
},
{
"epoch": 3.077835208269991,
"grad_norm": 0.30548402667045593,
"learning_rate": 1.5516976037000941e-06,
"loss": 0.5175,
"step": 634
},
{
"epoch": 3.082699908786865,
"grad_norm": 0.3132799565792084,
"learning_rate": 1.5363782324520033e-06,
"loss": 0.5207,
"step": 635
},
{
"epoch": 3.08756460930374,
"grad_norm": 0.3199690580368042,
"learning_rate": 1.5211211189134955e-06,
"loss": 0.5041,
"step": 636
},
{
"epoch": 3.092429309820614,
"grad_norm": 0.3378254473209381,
"learning_rate": 1.5059265373272574e-06,
"loss": 0.5122,
"step": 637
},
{
"epoch": 3.0972940103374884,
"grad_norm": 0.33055004477500916,
"learning_rate": 1.4907947608119866e-06,
"loss": 0.554,
"step": 638
},
{
"epoch": 3.102158710854363,
"grad_norm": 0.3463766574859619,
"learning_rate": 1.475726061357463e-06,
"loss": 0.5452,
"step": 639
},
{
"epoch": 3.1070234113712374,
"grad_norm": 0.3232378661632538,
"learning_rate": 1.4607207098196851e-06,
"loss": 0.5189,
"step": 640
},
{
"epoch": 3.111888111888112,
"grad_norm": 0.3677986264228821,
"learning_rate": 1.4457789759159813e-06,
"loss": 0.534,
"step": 641
},
{
"epoch": 3.1167528124049864,
"grad_norm": 0.31549063324928284,
"learning_rate": 1.430901128220174e-06,
"loss": 0.5109,
"step": 642
},
{
"epoch": 3.1216175129218606,
"grad_norm": 0.3554557263851166,
"learning_rate": 1.4160874341577447e-06,
"loss": 0.5379,
"step": 643
},
{
"epoch": 3.1264822134387353,
"grad_norm": 0.353786438703537,
"learning_rate": 1.4013381600010278e-06,
"loss": 0.5343,
"step": 644
},
{
"epoch": 3.1313469139556096,
"grad_norm": 0.35455793142318726,
"learning_rate": 1.3866535708644335e-06,
"loss": 0.5174,
"step": 645
},
{
"epoch": 3.136211614472484,
"grad_norm": 0.3173479735851288,
"learning_rate": 1.3720339306996666e-06,
"loss": 0.5396,
"step": 646
},
{
"epoch": 3.1410763149893586,
"grad_norm": 0.32542359828948975,
"learning_rate": 1.3574795022910014e-06,
"loss": 0.536,
"step": 647
},
{
"epoch": 3.145941015506233,
"grad_norm": 0.3286789655685425,
"learning_rate": 1.3429905472505344e-06,
"loss": 0.5357,
"step": 648
},
{
"epoch": 3.1508057160231075,
"grad_norm": 0.3170234262943268,
"learning_rate": 1.3285673260135073e-06,
"loss": 0.5213,
"step": 649
},
{
"epoch": 3.155670416539982,
"grad_norm": 0.3175816237926483,
"learning_rate": 1.314210097833607e-06,
"loss": 0.5085,
"step": 650
},
{
"epoch": 3.160535117056856,
"grad_norm": 0.3415636122226715,
"learning_rate": 1.2999191207783129e-06,
"loss": 0.5456,
"step": 651
},
{
"epoch": 3.165399817573731,
"grad_norm": 0.34509846568107605,
"learning_rate": 1.2856946517242608e-06,
"loss": 0.542,
"step": 652
},
{
"epoch": 3.170264518090605,
"grad_norm": 0.3505297899246216,
"learning_rate": 1.2715369463526173e-06,
"loss": 0.5385,
"step": 653
},
{
"epoch": 3.1751292186074793,
"grad_norm": 0.34468066692352295,
"learning_rate": 1.257446259144494e-06,
"loss": 0.518,
"step": 654
},
{
"epoch": 3.179993919124354,
"grad_norm": 0.3337211012840271,
"learning_rate": 1.2434228433763657e-06,
"loss": 0.5299,
"step": 655
},
{
"epoch": 3.1848586196412283,
"grad_norm": 0.35590267181396484,
"learning_rate": 1.2294669511155193e-06,
"loss": 0.5601,
"step": 656
},
{
"epoch": 3.1897233201581026,
"grad_norm": 0.336508572101593,
"learning_rate": 1.215578833215526e-06,
"loss": 0.5348,
"step": 657
},
{
"epoch": 3.1945880206749773,
"grad_norm": 0.33203017711639404,
"learning_rate": 1.201758739311728e-06,
"loss": 0.5286,
"step": 658
},
{
"epoch": 3.1994527211918515,
"grad_norm": 0.32940563559532166,
"learning_rate": 1.1880069178167586e-06,
"loss": 0.5244,
"step": 659
},
{
"epoch": 3.2043174217087262,
"grad_norm": 0.30095043778419495,
"learning_rate": 1.1743236159160654e-06,
"loss": 0.5285,
"step": 660
},
{
"epoch": 3.2091821222256005,
"grad_norm": 0.33882370591163635,
"learning_rate": 1.1607090795634802e-06,
"loss": 0.5495,
"step": 661
},
{
"epoch": 3.2140468227424748,
"grad_norm": 0.32042574882507324,
"learning_rate": 1.1471635534767877e-06,
"loss": 0.5394,
"step": 662
},
{
"epoch": 3.2189115232593495,
"grad_norm": 0.3295673727989197,
"learning_rate": 1.133687281133331e-06,
"loss": 0.5289,
"step": 663
},
{
"epoch": 3.2237762237762237,
"grad_norm": 0.3398403227329254,
"learning_rate": 1.1202805047656406e-06,
"loss": 0.5379,
"step": 664
},
{
"epoch": 3.228640924293098,
"grad_norm": 0.34369248151779175,
"learning_rate": 1.1069434653570633e-06,
"loss": 0.5185,
"step": 665
},
{
"epoch": 3.2335056248099727,
"grad_norm": 0.32993361353874207,
"learning_rate": 1.0936764026374547e-06,
"loss": 0.5359,
"step": 666
},
{
"epoch": 3.238370325326847,
"grad_norm": 0.32760289311408997,
"learning_rate": 1.0804795550788473e-06,
"loss": 0.5256,
"step": 667
},
{
"epoch": 3.2432350258437217,
"grad_norm": 0.3815072774887085,
"learning_rate": 1.0673531598911824e-06,
"loss": 0.5224,
"step": 668
},
{
"epoch": 3.248099726360596,
"grad_norm": 0.3164839744567871,
"learning_rate": 1.0542974530180327e-06,
"loss": 0.541,
"step": 669
},
{
"epoch": 3.2529644268774702,
"grad_norm": 0.34408512711524963,
"learning_rate": 1.0413126691323667e-06,
"loss": 0.5314,
"step": 670
},
{
"epoch": 3.257829127394345,
"grad_norm": 0.3566529154777527,
"learning_rate": 1.0283990416323336e-06,
"loss": 0.5287,
"step": 671
},
{
"epoch": 3.262693827911219,
"grad_norm": 0.34521618485450745,
"learning_rate": 1.0155568026370637e-06,
"loss": 0.5448,
"step": 672
},
{
"epoch": 3.2675585284280935,
"grad_norm": 0.33440014719963074,
"learning_rate": 1.0027861829824953e-06,
"loss": 0.5456,
"step": 673
},
{
"epoch": 3.272423228944968,
"grad_norm": 0.328668475151062,
"learning_rate": 9.900874122172294e-07,
"loss": 0.5244,
"step": 674
},
{
"epoch": 3.2772879294618424,
"grad_norm": 0.34540122747421265,
"learning_rate": 9.774607185984004e-07,
"loss": 0.5326,
"step": 675
},
{
"epoch": 3.282152629978717,
"grad_norm": 0.3642292022705078,
"learning_rate": 9.649063290875771e-07,
"loss": 0.5557,
"step": 676
},
{
"epoch": 3.2870173304955914,
"grad_norm": 0.3366372585296631,
"learning_rate": 9.524244693466773e-07,
"loss": 0.5378,
"step": 677
},
{
"epoch": 3.2918820310124657,
"grad_norm": 0.3133717477321625,
"learning_rate": 9.400153637339182e-07,
"loss": 0.5317,
"step": 678
},
{
"epoch": 3.2967467315293404,
"grad_norm": 0.3667609691619873,
"learning_rate": 9.276792352997782e-07,
"loss": 0.5653,
"step": 679
},
{
"epoch": 3.3016114320462147,
"grad_norm": 0.3069620728492737,
"learning_rate": 9.154163057829879e-07,
"loss": 0.5095,
"step": 680
},
{
"epoch": 3.306476132563089,
"grad_norm": 0.3310723900794983,
"learning_rate": 9.032267956065516e-07,
"loss": 0.5536,
"step": 681
},
{
"epoch": 3.3113408330799636,
"grad_norm": 0.33996835350990295,
"learning_rate": 8.911109238737748e-07,
"loss": 0.5116,
"step": 682
},
{
"epoch": 3.316205533596838,
"grad_norm": 0.36909931898117065,
"learning_rate": 8.790689083643328e-07,
"loss": 0.5683,
"step": 683
},
{
"epoch": 3.321070234113712,
"grad_norm": 0.33803001046180725,
"learning_rate": 8.671009655303531e-07,
"loss": 0.5487,
"step": 684
},
{
"epoch": 3.325934934630587,
"grad_norm": 0.27806195616722107,
"learning_rate": 8.552073104925296e-07,
"loss": 0.5182,
"step": 685
},
{
"epoch": 3.330799635147461,
"grad_norm": 0.3142351806163788,
"learning_rate": 8.433881570362484e-07,
"loss": 0.5225,
"step": 686
},
{
"epoch": 3.335664335664336,
"grad_norm": 0.32534122467041016,
"learning_rate": 8.316437176077491e-07,
"loss": 0.5458,
"step": 687
},
{
"epoch": 3.34052903618121,
"grad_norm": 0.3418058753013611,
"learning_rate": 8.199742033103091e-07,
"loss": 0.5264,
"step": 688
},
{
"epoch": 3.3453937366980844,
"grad_norm": 0.3124678134918213,
"learning_rate": 8.083798239004408e-07,
"loss": 0.5368,
"step": 689
},
{
"epoch": 3.350258437214959,
"grad_norm": 0.3275570273399353,
"learning_rate": 7.968607877841333e-07,
"loss": 0.5502,
"step": 690
},
{
"epoch": 3.3551231377318333,
"grad_norm": 0.30115506052970886,
"learning_rate": 7.854173020130906e-07,
"loss": 0.5256,
"step": 691
},
{
"epoch": 3.359987838248708,
"grad_norm": 0.3279692828655243,
"learning_rate": 7.740495722810271e-07,
"loss": 0.5302,
"step": 692
},
{
"epoch": 3.3648525387655823,
"grad_norm": 0.3461444079875946,
"learning_rate": 7.627578029199562e-07,
"loss": 0.5516,
"step": 693
},
{
"epoch": 3.3697172392824566,
"grad_norm": 0.317022442817688,
"learning_rate": 7.515421968965242e-07,
"loss": 0.53,
"step": 694
},
{
"epoch": 3.374581939799331,
"grad_norm": 0.33896568417549133,
"learning_rate": 7.404029558083653e-07,
"loss": 0.5272,
"step": 695
},
{
"epoch": 3.3794466403162056,
"grad_norm": 0.32342201471328735,
"learning_rate": 7.293402798804667e-07,
"loss": 0.5339,
"step": 696
},
{
"epoch": 3.38431134083308,
"grad_norm": 0.3396725356578827,
"learning_rate": 7.183543679615834e-07,
"loss": 0.5397,
"step": 697
},
{
"epoch": 3.3891760413499545,
"grad_norm": 0.33358529210090637,
"learning_rate": 7.074454175206524e-07,
"loss": 0.5598,
"step": 698
},
{
"epoch": 3.394040741866829,
"grad_norm": 0.3364230990409851,
"learning_rate": 6.966136246432492e-07,
"loss": 0.523,
"step": 699
},
{
"epoch": 3.398905442383703,
"grad_norm": 0.2987503111362457,
"learning_rate": 6.858591840280627e-07,
"loss": 0.5273,
"step": 700
},
{
"epoch": 3.4037701429005778,
"grad_norm": 0.32737722992897034,
"learning_rate": 6.751822889833926e-07,
"loss": 0.5442,
"step": 701
},
{
"epoch": 3.408634843417452,
"grad_norm": 0.32116642594337463,
"learning_rate": 6.645831314236817e-07,
"loss": 0.5339,
"step": 702
},
{
"epoch": 3.4134995439343268,
"grad_norm": 0.3150164484977722,
"learning_rate": 6.540619018660555e-07,
"loss": 0.5271,
"step": 703
},
{
"epoch": 3.418364244451201,
"grad_norm": 0.357954740524292,
"learning_rate": 6.436187894269086e-07,
"loss": 0.5434,
"step": 704
},
{
"epoch": 3.4232289449680753,
"grad_norm": 0.32282474637031555,
"learning_rate": 6.332539818184985e-07,
"loss": 0.5171,
"step": 705
},
{
"epoch": 3.42809364548495,
"grad_norm": 0.3020450174808502,
"learning_rate": 6.229676653455719e-07,
"loss": 0.5365,
"step": 706
},
{
"epoch": 3.4329583460018243,
"grad_norm": 0.32217666506767273,
"learning_rate": 6.127600249020216e-07,
"loss": 0.5423,
"step": 707
},
{
"epoch": 3.4378230465186985,
"grad_norm": 0.33344945311546326,
"learning_rate": 6.026312439675553e-07,
"loss": 0.5263,
"step": 708
},
{
"epoch": 3.4426877470355732,
"grad_norm": 0.3465080261230469,
"learning_rate": 5.925815046044026e-07,
"loss": 0.5298,
"step": 709
},
{
"epoch": 3.4475524475524475,
"grad_norm": 0.3112027049064636,
"learning_rate": 5.826109874540409e-07,
"loss": 0.5299,
"step": 710
},
{
"epoch": 3.4524171480693218,
"grad_norm": 0.3448041081428528,
"learning_rate": 5.727198717339511e-07,
"loss": 0.5635,
"step": 711
},
{
"epoch": 3.4572818485861965,
"grad_norm": 0.35311010479927063,
"learning_rate": 5.6290833523439e-07,
"loss": 0.542,
"step": 712
},
{
"epoch": 3.4621465491030707,
"grad_norm": 0.30904555320739746,
"learning_rate": 5.531765543152002e-07,
"loss": 0.5429,
"step": 713
},
{
"epoch": 3.4670112496199454,
"grad_norm": 0.31063464283943176,
"learning_rate": 5.435247039026398e-07,
"loss": 0.5208,
"step": 714
},
{
"epoch": 3.4718759501368197,
"grad_norm": 0.32749229669570923,
"learning_rate": 5.33952957486234e-07,
"loss": 0.5325,
"step": 715
},
{
"epoch": 3.476740650653694,
"grad_norm": 0.33042624592781067,
"learning_rate": 5.244614871156612e-07,
"loss": 0.5527,
"step": 716
},
{
"epoch": 3.4816053511705687,
"grad_norm": 0.32299965620040894,
"learning_rate": 5.150504633976572e-07,
"loss": 0.5216,
"step": 717
},
{
"epoch": 3.486470051687443,
"grad_norm": 0.32291823625564575,
"learning_rate": 5.057200554929509e-07,
"loss": 0.5541,
"step": 718
},
{
"epoch": 3.4913347522043177,
"grad_norm": 0.309826523065567,
"learning_rate": 4.964704311132224e-07,
"loss": 0.5231,
"step": 719
},
{
"epoch": 3.496199452721192,
"grad_norm": 0.31125494837760925,
"learning_rate": 4.873017565180871e-07,
"loss": 0.5276,
"step": 720
},
{
"epoch": 3.501064153238066,
"grad_norm": 0.316703736782074,
"learning_rate": 4.782141965121129e-07,
"loss": 0.5423,
"step": 721
},
{
"epoch": 3.5059288537549405,
"grad_norm": 0.32176005840301514,
"learning_rate": 4.6920791444184934e-07,
"loss": 0.5297,
"step": 722
},
{
"epoch": 3.510793554271815,
"grad_norm": 0.3091294765472412,
"learning_rate": 4.602830721928997e-07,
"loss": 0.5283,
"step": 723
},
{
"epoch": 3.5156582547886894,
"grad_norm": 0.322344034910202,
"learning_rate": 4.5143983018700485e-07,
"loss": 0.5421,
"step": 724
},
{
"epoch": 3.520522955305564,
"grad_norm": 0.35230642557144165,
"learning_rate": 4.4267834737916295e-07,
"loss": 0.5531,
"step": 725
},
{
"epoch": 3.5253876558224384,
"grad_norm": 0.3391262888908386,
"learning_rate": 4.33998781254773e-07,
"loss": 0.5604,
"step": 726
},
{
"epoch": 3.5302523563393127,
"grad_norm": 0.31481173634529114,
"learning_rate": 4.2540128782679934e-07,
"loss": 0.5289,
"step": 727
},
{
"epoch": 3.5351170568561874,
"grad_norm": 0.32447385787963867,
"learning_rate": 4.1688602163297564e-07,
"loss": 0.5299,
"step": 728
},
{
"epoch": 3.5399817573730616,
"grad_norm": 0.3530765771865845,
"learning_rate": 4.0845313573301736e-07,
"loss": 0.5464,
"step": 729
},
{
"epoch": 3.5448464578899364,
"grad_norm": 0.30712568759918213,
"learning_rate": 4.001027817058789e-07,
"loss": 0.5242,
"step": 730
},
{
"epoch": 3.5497111584068106,
"grad_norm": 0.3143819272518158,
"learning_rate": 3.9183510964702463e-07,
"loss": 0.5273,
"step": 731
},
{
"epoch": 3.554575858923685,
"grad_norm": 0.3388993740081787,
"learning_rate": 3.836502681657289e-07,
"loss": 0.5335,
"step": 732
},
{
"epoch": 3.5594405594405596,
"grad_norm": 0.3553450405597687,
"learning_rate": 3.755484043824131e-07,
"loss": 0.5516,
"step": 733
},
{
"epoch": 3.564305259957434,
"grad_norm": 0.3086324632167816,
"learning_rate": 3.675296639259912e-07,
"loss": 0.5445,
"step": 734
},
{
"epoch": 3.5691699604743086,
"grad_norm": 0.3342626392841339,
"learning_rate": 3.595941909312595e-07,
"loss": 0.5162,
"step": 735
},
{
"epoch": 3.574034660991183,
"grad_norm": 0.32068684697151184,
"learning_rate": 3.517421280363004e-07,
"loss": 0.5401,
"step": 736
},
{
"epoch": 3.578899361508057,
"grad_norm": 0.33277034759521484,
"learning_rate": 3.439736163799251e-07,
"loss": 0.5149,
"step": 737
},
{
"epoch": 3.5837640620249314,
"grad_norm": 0.3014756143093109,
"learning_rate": 3.362887955991301e-07,
"loss": 0.512,
"step": 738
},
{
"epoch": 3.588628762541806,
"grad_norm": 0.3068234324455261,
"learning_rate": 3.2868780382658895e-07,
"loss": 0.5209,
"step": 739
},
{
"epoch": 3.5934934630586803,
"grad_norm": 0.3191012740135193,
"learning_rate": 3.2117077768817395e-07,
"loss": 0.5597,
"step": 740
},
{
"epoch": 3.598358163575555,
"grad_norm": 0.32741162180900574,
"learning_rate": 3.1373785230049356e-07,
"loss": 0.5515,
"step": 741
},
{
"epoch": 3.6032228640924293,
"grad_norm": 0.342742919921875,
"learning_rate": 3.0638916126846885e-07,
"loss": 0.5446,
"step": 742
},
{
"epoch": 3.6080875646093036,
"grad_norm": 0.31945279240608215,
"learning_rate": 2.991248366829291e-07,
"loss": 0.5426,
"step": 743
},
{
"epoch": 3.6129522651261783,
"grad_norm": 0.3091720640659332,
"learning_rate": 2.91945009118238e-07,
"loss": 0.5357,
"step": 744
},
{
"epoch": 3.6178169656430526,
"grad_norm": 0.32953116297721863,
"learning_rate": 2.848498076299483e-07,
"loss": 0.5392,
"step": 745
},
{
"epoch": 3.6226816661599273,
"grad_norm": 0.3301558494567871,
"learning_rate": 2.7783935975247867e-07,
"loss": 0.5266,
"step": 746
},
{
"epoch": 3.6275463666768015,
"grad_norm": 0.31618431210517883,
"learning_rate": 2.7091379149682683e-07,
"loss": 0.5208,
"step": 747
},
{
"epoch": 3.632411067193676,
"grad_norm": 0.3258410394191742,
"learning_rate": 2.6407322734829763e-07,
"loss": 0.5262,
"step": 748
},
{
"epoch": 3.63727576771055,
"grad_norm": 0.2992360293865204,
"learning_rate": 2.573177902642726e-07,
"loss": 0.5272,
"step": 749
},
{
"epoch": 3.6421404682274248,
"grad_norm": 0.3194170892238617,
"learning_rate": 2.506476016719922e-07,
"loss": 0.5227,
"step": 750
},
{
"epoch": 3.647005168744299,
"grad_norm": 0.3228798806667328,
"learning_rate": 2.440627814663804e-07,
"loss": 0.5387,
"step": 751
},
{
"epoch": 3.6518698692611737,
"grad_norm": 0.3276216685771942,
"learning_rate": 2.3756344800788421e-07,
"loss": 0.535,
"step": 752
},
{
"epoch": 3.656734569778048,
"grad_norm": 0.31080737709999084,
"learning_rate": 2.3114971812034981e-07,
"loss": 0.5402,
"step": 753
},
{
"epoch": 3.6615992702949223,
"grad_norm": 0.3608711361885071,
"learning_rate": 2.2482170708892083e-07,
"loss": 0.5636,
"step": 754
},
{
"epoch": 3.666463970811797,
"grad_norm": 0.31172358989715576,
"learning_rate": 2.1857952865796616e-07,
"loss": 0.5526,
"step": 755
},
{
"epoch": 3.6713286713286712,
"grad_norm": 0.3504408299922943,
"learning_rate": 2.124232950290367e-07,
"loss": 0.541,
"step": 756
},
{
"epoch": 3.676193371845546,
"grad_norm": 0.3059706687927246,
"learning_rate": 2.0635311685884675e-07,
"loss": 0.5225,
"step": 757
},
{
"epoch": 3.68105807236242,
"grad_norm": 0.30679747462272644,
"learning_rate": 2.0036910325728521e-07,
"loss": 0.5287,
"step": 758
},
{
"epoch": 3.6859227728792945,
"grad_norm": 0.3577539920806885,
"learning_rate": 1.9447136178545766e-07,
"loss": 0.5645,
"step": 759
},
{
"epoch": 3.690787473396169,
"grad_norm": 0.3062092661857605,
"learning_rate": 1.8865999845374794e-07,
"loss": 0.5346,
"step": 760
},
{
"epoch": 3.6956521739130435,
"grad_norm": 0.31235286593437195,
"learning_rate": 1.8293511771991624e-07,
"loss": 0.5485,
"step": 761
},
{
"epoch": 3.700516874429918,
"grad_norm": 0.3473973274230957,
"learning_rate": 1.7729682248721848e-07,
"loss": 0.5169,
"step": 762
},
{
"epoch": 3.7053815749467924,
"grad_norm": 0.31875768303871155,
"learning_rate": 1.7174521410256162e-07,
"loss": 0.5264,
"step": 763
},
{
"epoch": 3.7102462754636667,
"grad_norm": 0.345159649848938,
"learning_rate": 1.6628039235467686e-07,
"loss": 0.5253,
"step": 764
},
{
"epoch": 3.715110975980541,
"grad_norm": 0.31550589203834534,
"learning_rate": 1.6090245547232707e-07,
"loss": 0.5347,
"step": 765
},
{
"epoch": 3.7199756764974157,
"grad_norm": 0.31688812375068665,
"learning_rate": 1.5561150012254446e-07,
"loss": 0.5437,
"step": 766
},
{
"epoch": 3.72484037701429,
"grad_norm": 0.33005401492118835,
"learning_rate": 1.5040762140888843e-07,
"loss": 0.5182,
"step": 767
},
{
"epoch": 3.7297050775311646,
"grad_norm": 0.32252198457717896,
"learning_rate": 1.4529091286973994e-07,
"loss": 0.5343,
"step": 768
},
{
"epoch": 3.734569778048039,
"grad_norm": 0.34058618545532227,
"learning_rate": 1.402614664766172e-07,
"loss": 0.5682,
"step": 769
},
{
"epoch": 3.739434478564913,
"grad_norm": 0.3218619227409363,
"learning_rate": 1.353193726325247e-07,
"loss": 0.5523,
"step": 770
},
{
"epoch": 3.744299179081788,
"grad_norm": 0.338169127702713,
"learning_rate": 1.3046472017032685e-07,
"loss": 0.5426,
"step": 771
},
{
"epoch": 3.749163879598662,
"grad_norm": 0.3439265191555023,
"learning_rate": 1.2569759635115086e-07,
"loss": 0.5586,
"step": 772
},
{
"epoch": 3.754028580115537,
"grad_norm": 0.3248153626918793,
"learning_rate": 1.210180868628219e-07,
"loss": 0.5317,
"step": 773
},
{
"epoch": 3.758893280632411,
"grad_norm": 0.31938469409942627,
"learning_rate": 1.1642627581831767e-07,
"loss": 0.5214,
"step": 774
},
{
"epoch": 3.7637579811492854,
"grad_norm": 0.30051711201667786,
"learning_rate": 1.1192224575425848e-07,
"loss": 0.5465,
"step": 775
},
{
"epoch": 3.7686226816661597,
"grad_norm": 0.2911449372768402,
"learning_rate": 1.0750607762942622e-07,
"loss": 0.537,
"step": 776
},
{
"epoch": 3.7734873821830344,
"grad_norm": 0.2878769338130951,
"learning_rate": 1.0317785082330555e-07,
"loss": 0.5352,
"step": 777
},
{
"epoch": 3.7783520826999086,
"grad_norm": 0.3339023292064667,
"learning_rate": 9.89376431346606e-08,
"loss": 0.5606,
"step": 778
},
{
"epoch": 3.7832167832167833,
"grad_norm": 0.31095194816589355,
"learning_rate": 9.478553078013042e-08,
"loss": 0.5248,
"step": 779
},
{
"epoch": 3.7880814837336576,
"grad_norm": 0.31098252534866333,
"learning_rate": 9.072158839286748e-08,
"loss": 0.532,
"step": 780
},
{
"epoch": 3.792946184250532,
"grad_norm": 0.31499919295310974,
"learning_rate": 8.674588902118919e-08,
"loss": 0.5193,
"step": 781
},
{
"epoch": 3.7978108847674066,
"grad_norm": 0.3318183422088623,
"learning_rate": 8.285850412726837e-08,
"loss": 0.5308,
"step": 782
},
{
"epoch": 3.802675585284281,
"grad_norm": 0.35332199931144714,
"learning_rate": 7.905950358584768e-08,
"loss": 0.5495,
"step": 783
},
{
"epoch": 3.8075402858011556,
"grad_norm": 0.34943094849586487,
"learning_rate": 7.534895568298395e-08,
"loss": 0.5416,
"step": 784
},
{
"epoch": 3.81240498631803,
"grad_norm": 0.31405407190322876,
"learning_rate": 7.172692711482022e-08,
"loss": 0.5565,
"step": 785
},
{
"epoch": 3.817269686834904,
"grad_norm": 0.35813677310943604,
"learning_rate": 6.819348298638839e-08,
"loss": 0.5618,
"step": 786
},
{
"epoch": 3.822134387351779,
"grad_norm": 0.30600205063819885,
"learning_rate": 6.474868681043578e-08,
"loss": 0.5264,
"step": 787
},
{
"epoch": 3.826999087868653,
"grad_norm": 0.3123580813407898,
"learning_rate": 6.13926005062876e-08,
"loss": 0.5209,
"step": 788
},
{
"epoch": 3.8318637883855278,
"grad_norm": 0.32482513785362244,
"learning_rate": 5.8125284398730666e-08,
"loss": 0.5192,
"step": 789
},
{
"epoch": 3.836728488902402,
"grad_norm": 0.34198108315467834,
"learning_rate": 5.4946797216931524e-08,
"loss": 0.5449,
"step": 790
},
{
"epoch": 3.8415931894192763,
"grad_norm": 0.31200042366981506,
"learning_rate": 5.185719609337836e-08,
"loss": 0.5111,
"step": 791
},
{
"epoch": 3.8464578899361506,
"grad_norm": 0.3532400131225586,
"learning_rate": 4.885653656285627e-08,
"loss": 0.5233,
"step": 792
},
{
"epoch": 3.8513225904530253,
"grad_norm": 0.32765015959739685,
"learning_rate": 4.5944872561448084e-08,
"loss": 0.5606,
"step": 793
},
{
"epoch": 3.8561872909698995,
"grad_norm": 0.3308298885822296,
"learning_rate": 4.3122256425563444e-08,
"loss": 0.5384,
"step": 794
},
{
"epoch": 3.8610519914867742,
"grad_norm": 0.33672013878822327,
"learning_rate": 4.038873889100237e-08,
"loss": 0.5516,
"step": 795
},
{
"epoch": 3.8659166920036485,
"grad_norm": 0.33695557713508606,
"learning_rate": 3.77443690920376e-08,
"loss": 0.5689,
"step": 796
},
{
"epoch": 3.870781392520523,
"grad_norm": 0.2859351336956024,
"learning_rate": 3.518919456053649e-08,
"loss": 0.5302,
"step": 797
},
{
"epoch": 3.8756460930373975,
"grad_norm": 0.3394811749458313,
"learning_rate": 3.2723261225102164e-08,
"loss": 0.5432,
"step": 798
},
{
"epoch": 3.8805107935542718,
"grad_norm": 0.3551604449748993,
"learning_rate": 3.034661341025258e-08,
"loss": 0.5324,
"step": 799
},
{
"epoch": 3.8853754940711465,
"grad_norm": 0.31553763151168823,
"learning_rate": 2.8059293835620006e-08,
"loss": 0.5305,
"step": 800
},
{
"epoch": 3.8902401945880207,
"grad_norm": 0.32441508769989014,
"learning_rate": 2.5861343615184997e-08,
"loss": 0.5415,
"step": 801
},
{
"epoch": 3.895104895104895,
"grad_norm": 0.31447526812553406,
"learning_rate": 2.3752802256536423e-08,
"loss": 0.5411,
"step": 802
},
{
"epoch": 3.8999695956217693,
"grad_norm": 0.3224940299987793,
"learning_rate": 2.173370766016314e-08,
"loss": 0.5197,
"step": 803
},
{
"epoch": 3.904834296138644,
"grad_norm": 0.29226142168045044,
"learning_rate": 1.98040961187701e-08,
"loss": 0.5182,
"step": 804
},
{
"epoch": 3.9096989966555182,
"grad_norm": 0.32759496569633484,
"learning_rate": 1.7964002316628316e-08,
"loss": 0.5395,
"step": 805
},
{
"epoch": 3.914563697172393,
"grad_norm": 0.34152671694755554,
"learning_rate": 1.6213459328950355e-08,
"loss": 0.5614,
"step": 806
},
{
"epoch": 3.919428397689267,
"grad_norm": 0.30443131923675537,
"learning_rate": 1.4552498621295264e-08,
"loss": 0.5167,
"step": 807
},
{
"epoch": 3.9242930982061415,
"grad_norm": 0.2858695089817047,
"learning_rate": 1.2981150049004021e-08,
"loss": 0.5277,
"step": 808
},
{
"epoch": 3.929157798723016,
"grad_norm": 0.3150896728038788,
"learning_rate": 1.1499441856663296e-08,
"loss": 0.5496,
"step": 809
},
{
"epoch": 3.9340224992398904,
"grad_norm": 0.3497582972049713,
"learning_rate": 1.0107400677596413e-08,
"loss": 0.5611,
"step": 810
},
{
"epoch": 3.938887199756765,
"grad_norm": 0.3036641776561737,
"learning_rate": 8.805051533384846e-09,
"loss": 0.5345,
"step": 811
},
{
"epoch": 3.9437519002736394,
"grad_norm": 0.31381335854530334,
"learning_rate": 7.59241783341913e-09,
"loss": 0.527,
"step": 812
},
{
"epoch": 3.9486166007905137,
"grad_norm": 0.31904760003089905,
"learning_rate": 6.469521374477539e-09,
"loss": 0.53,
"step": 813
},
{
"epoch": 3.9534813013073884,
"grad_norm": 0.3499225378036499,
"learning_rate": 5.436382340335833e-09,
"loss": 0.5377,
"step": 814
},
{
"epoch": 3.9583460018242627,
"grad_norm": 0.2925429344177246,
"learning_rate": 4.493019301401447e-09,
"loss": 0.5185,
"step": 815
},
{
"epoch": 3.9632107023411374,
"grad_norm": 0.335538774728775,
"learning_rate": 3.6394492143820847e-09,
"loss": 0.5387,
"step": 816
},
{
"epoch": 3.9680754028580116,
"grad_norm": 0.2967749834060669,
"learning_rate": 2.875687421980966e-09,
"loss": 0.5394,
"step": 817
},
{
"epoch": 3.972940103374886,
"grad_norm": 0.3312070369720459,
"learning_rate": 2.201747652618713e-09,
"loss": 0.5375,
"step": 818
},
{
"epoch": 3.97780480389176,
"grad_norm": 0.3883361518383026,
"learning_rate": 1.6176420201902132e-09,
"loss": 0.5434,
"step": 819
},
{
"epoch": 3.982669504408635,
"grad_norm": 0.33711862564086914,
"learning_rate": 1.1233810238425735e-09,
"loss": 0.5432,
"step": 820
},
{
"epoch": 3.987534204925509,
"grad_norm": 0.3115112781524658,
"learning_rate": 7.189735477913795e-10,
"loss": 0.5247,
"step": 821
},
{
"epoch": 3.992398905442384,
"grad_norm": 0.307807981967926,
"learning_rate": 4.0442686115582665e-10,
"loss": 0.5227,
"step": 822
},
{
"epoch": 3.997263605959258,
"grad_norm": 0.3572421967983246,
"learning_rate": 1.797466178327101e-10,
"loss": 0.538,
"step": 823
},
{
"epoch": 4.0,
"grad_norm": 0.43410807847976685,
"learning_rate": 4.4936856390398465e-11,
"loss": 0.5144,
"step": 824
},
{
"epoch": 4.0,
"step": 824,
"total_flos": 2.006101437057283e+19,
"train_loss": 0.5864963867131946,
"train_runtime": 153073.7159,
"train_samples_per_second": 0.687,
"train_steps_per_second": 0.005
}
],
"logging_steps": 1,
"max_steps": 824,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.006101437057283e+19,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}