Files
wanna-pilot/checkpoint-7813/trainer_state.json
ModelHub XC 9d63e58716 初始化项目,由ModelHub XC社区提供模型
Model: bbunzeck/wanna-pilot
Source: Original Platform
2026-07-30 01:37:18 +08:00

5510 lines
134 KiB
JSON

{
"best_global_step": 7813,
"best_metric": 2.000473737716675,
"best_model_checkpoint": "/mnt/myscratch/model/checkpoint-7813",
"epoch": 1.0,
"eval_steps": 500,
"global_step": 7813,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00128,
"grad_norm": 5.885704040527344,
"learning_rate": 1.3499999999999998e-05,
"loss": 6.605717468261719,
"step": 10
},
{
"epoch": 0.00256,
"grad_norm": 4.186084747314453,
"learning_rate": 2.8499999999999998e-05,
"loss": 5.801454544067383,
"step": 20
},
{
"epoch": 0.00384,
"grad_norm": 2.714238166809082,
"learning_rate": 4.3499999999999993e-05,
"loss": 4.964897155761719,
"step": 30
},
{
"epoch": 0.00512,
"grad_norm": 2.2332475185394287,
"learning_rate": 5.85e-05,
"loss": 4.384710693359375,
"step": 40
},
{
"epoch": 0.0064,
"grad_norm": 2.1372387409210205,
"learning_rate": 7.35e-05,
"loss": 3.9424991607666016,
"step": 50
},
{
"epoch": 0.00768,
"grad_norm": 2.8777883052825928,
"learning_rate": 8.849999999999998e-05,
"loss": 3.6049922943115233,
"step": 60
},
{
"epoch": 0.00896,
"grad_norm": 2.3045167922973633,
"learning_rate": 0.00010349999999999998,
"loss": 3.2790637969970704,
"step": 70
},
{
"epoch": 0.01024,
"grad_norm": 3.263000011444092,
"learning_rate": 0.0001185,
"loss": 3.0744760513305662,
"step": 80
},
{
"epoch": 0.01152,
"grad_norm": 2.0662527084350586,
"learning_rate": 0.0001335,
"loss": 2.8297033309936523,
"step": 90
},
{
"epoch": 0.0128,
"grad_norm": 2.5418498516082764,
"learning_rate": 0.00014849999999999998,
"loss": 2.659013366699219,
"step": 100
},
{
"epoch": 0.01408,
"grad_norm": 1.6310275793075562,
"learning_rate": 0.0001635,
"loss": 2.49525146484375,
"step": 110
},
{
"epoch": 0.01536,
"grad_norm": 2.0985260009765625,
"learning_rate": 0.00017849999999999997,
"loss": 2.42862491607666,
"step": 120
},
{
"epoch": 0.01664,
"grad_norm": 1.5091782808303833,
"learning_rate": 0.0001935,
"loss": 2.3671804428100587,
"step": 130
},
{
"epoch": 0.01792,
"grad_norm": 1.6089136600494385,
"learning_rate": 0.00020849999999999997,
"loss": 2.261024284362793,
"step": 140
},
{
"epoch": 0.0192,
"grad_norm": 1.0941425561904907,
"learning_rate": 0.00022349999999999998,
"loss": 2.221732330322266,
"step": 150
},
{
"epoch": 0.02048,
"grad_norm": 1.2431774139404297,
"learning_rate": 0.0002385,
"loss": 2.2301746368408204,
"step": 160
},
{
"epoch": 0.02176,
"grad_norm": 1.1251002550125122,
"learning_rate": 0.0002535,
"loss": 2.204315948486328,
"step": 170
},
{
"epoch": 0.02304,
"grad_norm": 1.195177435874939,
"learning_rate": 0.00026849999999999997,
"loss": 2.170834541320801,
"step": 180
},
{
"epoch": 0.02432,
"grad_norm": 1.409263014793396,
"learning_rate": 0.00028349999999999995,
"loss": 2.1824182510375976,
"step": 190
},
{
"epoch": 0.0256,
"grad_norm": 1.04000985622406,
"learning_rate": 0.0002985,
"loss": 2.1573196411132813,
"step": 200
},
{
"epoch": 0.02688,
"grad_norm": 1.1753369569778442,
"learning_rate": 0.0002999999901272834,
"loss": 2.1714319229125976,
"step": 210
},
{
"epoch": 0.02816,
"grad_norm": 1.028172492980957,
"learning_rate": 0.0002999999559993758,
"loss": 2.1164165496826173,
"step": 220
},
{
"epoch": 0.02944,
"grad_norm": 1.2896201610565186,
"learning_rate": 0.00029999989749439744,
"loss": 2.112924575805664,
"step": 230
},
{
"epoch": 0.03072,
"grad_norm": 1.0579850673675537,
"learning_rate": 0.00029999981461235783,
"loss": 2.114685821533203,
"step": 240
},
{
"epoch": 0.032,
"grad_norm": 0.8515546321868896,
"learning_rate": 0.00029999970735327037,
"loss": 2.0852130889892577,
"step": 250
},
{
"epoch": 0.03328,
"grad_norm": 1.0059428215026855,
"learning_rate": 0.0002999995757171525,
"loss": 2.0961055755615234,
"step": 260
},
{
"epoch": 0.03456,
"grad_norm": 0.8511396050453186,
"learning_rate": 0.0002999994197040256,
"loss": 2.0771358489990233,
"step": 270
},
{
"epoch": 0.03584,
"grad_norm": 0.9550066590309143,
"learning_rate": 0.0002999992393139151,
"loss": 2.070794677734375,
"step": 280
},
{
"epoch": 0.03712,
"grad_norm": 0.9620731472969055,
"learning_rate": 0.0002999990345468503,
"loss": 2.059033203125,
"step": 290
},
{
"epoch": 0.0384,
"grad_norm": 1.1329026222229004,
"learning_rate": 0.0002999988054028644,
"loss": 2.0862873077392576,
"step": 300
},
{
"epoch": 0.03968,
"grad_norm": 0.8112749457359314,
"learning_rate": 0.00029999855188199474,
"loss": 2.057857704162598,
"step": 310
},
{
"epoch": 0.04096,
"grad_norm": 0.9323590397834778,
"learning_rate": 0.0002999982739842824,
"loss": 2.0480510711669924,
"step": 320
},
{
"epoch": 0.04224,
"grad_norm": 0.8546434640884399,
"learning_rate": 0.0002999979717097727,
"loss": 2.067165756225586,
"step": 330
},
{
"epoch": 0.04352,
"grad_norm": 1.002989411354065,
"learning_rate": 0.00029999764505851456,
"loss": 2.0613639831542967,
"step": 340
},
{
"epoch": 0.0448,
"grad_norm": 0.8182193636894226,
"learning_rate": 0.00029999729403056127,
"loss": 2.0515058517456053,
"step": 350
},
{
"epoch": 0.04608,
"grad_norm": 0.7834963202476501,
"learning_rate": 0.00029999691862596975,
"loss": 2.046108436584473,
"step": 360
},
{
"epoch": 0.04736,
"grad_norm": 0.9317572712898254,
"learning_rate": 0.000299996518844801,
"loss": 2.0235712051391603,
"step": 370
},
{
"epoch": 0.04864,
"grad_norm": 0.8493145108222961,
"learning_rate": 0.0002999960946871201,
"loss": 2.0325557708740236,
"step": 380
},
{
"epoch": 0.04992,
"grad_norm": 0.922864556312561,
"learning_rate": 0.0002999956461529959,
"loss": 2.0653444290161134,
"step": 390
},
{
"epoch": 0.0512,
"grad_norm": 1.0530719757080078,
"learning_rate": 0.0002999951732425013,
"loss": 2.0602413177490235,
"step": 400
},
{
"epoch": 0.05248,
"grad_norm": 0.8999637961387634,
"learning_rate": 0.0002999946759557132,
"loss": 2.0521976470947267,
"step": 410
},
{
"epoch": 0.05376,
"grad_norm": 0.8567571640014648,
"learning_rate": 0.00029999415429271233,
"loss": 2.042643928527832,
"step": 420
},
{
"epoch": 0.05504,
"grad_norm": 0.9853494167327881,
"learning_rate": 0.00029999360825358354,
"loss": 2.043178939819336,
"step": 430
},
{
"epoch": 0.05632,
"grad_norm": 0.9435704946517944,
"learning_rate": 0.00029999303783841553,
"loss": 2.0274379730224608,
"step": 440
},
{
"epoch": 0.0576,
"grad_norm": 0.9081392288208008,
"learning_rate": 0.00029999244304730106,
"loss": 2.0362335205078126,
"step": 450
},
{
"epoch": 0.05888,
"grad_norm": 0.7286208868026733,
"learning_rate": 0.0002999918238803367,
"loss": 2.0380199432373045,
"step": 460
},
{
"epoch": 0.06016,
"grad_norm": 0.8928585648536682,
"learning_rate": 0.00029999118033762314,
"loss": 2.031109428405762,
"step": 470
},
{
"epoch": 0.06144,
"grad_norm": 0.8266701698303223,
"learning_rate": 0.00029999051241926493,
"loss": 2.0246667861938477,
"step": 480
},
{
"epoch": 0.06272,
"grad_norm": 0.7270641922950745,
"learning_rate": 0.00029998982012537063,
"loss": 2.042285346984863,
"step": 490
},
{
"epoch": 0.064,
"grad_norm": 0.7794978022575378,
"learning_rate": 0.0002999891034560528,
"loss": 2.0380592346191406,
"step": 500
},
{
"epoch": 0.06528,
"grad_norm": 0.8687776327133179,
"learning_rate": 0.00029998836241142786,
"loss": 2.0340011596679686,
"step": 510
},
{
"epoch": 0.06656,
"grad_norm": 0.8388617634773254,
"learning_rate": 0.0002999875969916162,
"loss": 2.035272979736328,
"step": 520
},
{
"epoch": 0.06784,
"grad_norm": 0.8601788878440857,
"learning_rate": 0.00029998680719674225,
"loss": 2.0376710891723633,
"step": 530
},
{
"epoch": 0.06912,
"grad_norm": 0.872636616230011,
"learning_rate": 0.00029998599302693434,
"loss": 2.0254486083984373,
"step": 540
},
{
"epoch": 0.0704,
"grad_norm": 0.7168036699295044,
"learning_rate": 0.00029998515448232484,
"loss": 2.044236755371094,
"step": 550
},
{
"epoch": 0.07168,
"grad_norm": 0.7704122066497803,
"learning_rate": 0.00029998429156304995,
"loss": 2.0272247314453127,
"step": 560
},
{
"epoch": 0.07296,
"grad_norm": 0.7757558822631836,
"learning_rate": 0.00029998340426925,
"loss": 2.0317691802978515,
"step": 570
},
{
"epoch": 0.07424,
"grad_norm": 0.9410697817802429,
"learning_rate": 0.0002999824926010691,
"loss": 2.0405961990356447,
"step": 580
},
{
"epoch": 0.07552,
"grad_norm": 0.7462645173072815,
"learning_rate": 0.0002999815565586555,
"loss": 2.0344850540161135,
"step": 590
},
{
"epoch": 0.0768,
"grad_norm": 0.8859974145889282,
"learning_rate": 0.00029998059614216123,
"loss": 2.0125873565673826,
"step": 600
},
{
"epoch": 0.07808,
"grad_norm": 0.724909782409668,
"learning_rate": 0.00029997961135174236,
"loss": 2.03349609375,
"step": 610
},
{
"epoch": 0.07936,
"grad_norm": 0.7383590936660767,
"learning_rate": 0.000299978602187559,
"loss": 2.033146095275879,
"step": 620
},
{
"epoch": 0.08064,
"grad_norm": 1.3143054246902466,
"learning_rate": 0.00029997756864977515,
"loss": 2.0456708908081054,
"step": 630
},
{
"epoch": 0.08192,
"grad_norm": 0.9198774099349976,
"learning_rate": 0.0002999765107385588,
"loss": 2.029145050048828,
"step": 640
},
{
"epoch": 0.0832,
"grad_norm": 0.7761960625648499,
"learning_rate": 0.00029997542845408174,
"loss": 2.040088653564453,
"step": 650
},
{
"epoch": 0.08448,
"grad_norm": 0.7530298829078674,
"learning_rate": 0.00029997432179652,
"loss": 2.0398796081542967,
"step": 660
},
{
"epoch": 0.08576,
"grad_norm": 0.7231208682060242,
"learning_rate": 0.0002999731907660533,
"loss": 2.029999542236328,
"step": 670
},
{
"epoch": 0.08704,
"grad_norm": 0.7728089094161987,
"learning_rate": 0.00029997203536286565,
"loss": 2.036482810974121,
"step": 680
},
{
"epoch": 0.08832,
"grad_norm": 0.607453465461731,
"learning_rate": 0.0002999708555871446,
"loss": 2.017901611328125,
"step": 690
},
{
"epoch": 0.0896,
"grad_norm": 0.779467761516571,
"learning_rate": 0.00029996965143908196,
"loss": 2.030159759521484,
"step": 700
},
{
"epoch": 0.09088,
"grad_norm": 0.7885509133338928,
"learning_rate": 0.00029996842291887344,
"loss": 2.0309608459472654,
"step": 710
},
{
"epoch": 0.09216,
"grad_norm": 0.7816950678825378,
"learning_rate": 0.0002999671700267187,
"loss": 1.9948287963867188,
"step": 720
},
{
"epoch": 0.09344,
"grad_norm": 0.806575357913971,
"learning_rate": 0.00029996589276282137,
"loss": 2.044095039367676,
"step": 730
},
{
"epoch": 0.09472,
"grad_norm": 0.7333192825317383,
"learning_rate": 0.0002999645911273889,
"loss": 2.038722610473633,
"step": 740
},
{
"epoch": 0.096,
"grad_norm": 0.8800010085105896,
"learning_rate": 0.00029996326512063295,
"loss": 2.035386085510254,
"step": 750
},
{
"epoch": 0.09728,
"grad_norm": 0.7811752557754517,
"learning_rate": 0.000299961914742769,
"loss": 2.004504585266113,
"step": 760
},
{
"epoch": 0.09856,
"grad_norm": 0.7882314920425415,
"learning_rate": 0.00029996053999401646,
"loss": 2.0211507797241213,
"step": 770
},
{
"epoch": 0.09984,
"grad_norm": 0.698653519153595,
"learning_rate": 0.0002999591408745988,
"loss": 2.0169116973876955,
"step": 780
},
{
"epoch": 0.10112,
"grad_norm": 0.6910223960876465,
"learning_rate": 0.00029995771738474335,
"loss": 2.002583694458008,
"step": 790
},
{
"epoch": 0.1024,
"grad_norm": 1.0781288146972656,
"learning_rate": 0.0002999562695246814,
"loss": 2.040352630615234,
"step": 800
},
{
"epoch": 0.10368,
"grad_norm": 0.7604212760925293,
"learning_rate": 0.0002999547972946484,
"loss": 2.012590789794922,
"step": 810
},
{
"epoch": 0.10496,
"grad_norm": 0.8433690071105957,
"learning_rate": 0.0002999533006948835,
"loss": 2.0214279174804686,
"step": 820
},
{
"epoch": 0.10624,
"grad_norm": 0.7420530319213867,
"learning_rate": 0.00029995177972562984,
"loss": 2.01849422454834,
"step": 830
},
{
"epoch": 0.10752,
"grad_norm": 0.8684757351875305,
"learning_rate": 0.00029995023438713475,
"loss": 2.0245651245117187,
"step": 840
},
{
"epoch": 0.1088,
"grad_norm": 0.8758244514465332,
"learning_rate": 0.0002999486646796493,
"loss": 2.0273792266845705,
"step": 850
},
{
"epoch": 0.11008,
"grad_norm": 0.7705071568489075,
"learning_rate": 0.0002999470706034286,
"loss": 2.003208351135254,
"step": 860
},
{
"epoch": 0.11136,
"grad_norm": 0.7473205327987671,
"learning_rate": 0.0002999454521587317,
"loss": 2.0413352966308596,
"step": 870
},
{
"epoch": 0.11264,
"grad_norm": 0.8020169138908386,
"learning_rate": 0.0002999438093458217,
"loss": 2.043915939331055,
"step": 880
},
{
"epoch": 0.11392,
"grad_norm": 0.8231315016746521,
"learning_rate": 0.00029994214216496544,
"loss": 2.0179935455322267,
"step": 890
},
{
"epoch": 0.1152,
"grad_norm": 0.8677017092704773,
"learning_rate": 0.00029994045061643393,
"loss": 2.0080083847045898,
"step": 900
},
{
"epoch": 0.11648,
"grad_norm": 0.7265374660491943,
"learning_rate": 0.0002999387347005021,
"loss": 2.0255220413208006,
"step": 910
},
{
"epoch": 0.11776,
"grad_norm": 0.7804266810417175,
"learning_rate": 0.0002999369944174487,
"loss": 2.0323333740234375,
"step": 920
},
{
"epoch": 0.11904,
"grad_norm": 1.909376859664917,
"learning_rate": 0.00029993522976755674,
"loss": 2.0403774261474608,
"step": 930
},
{
"epoch": 0.12032,
"grad_norm": 0.7820098996162415,
"learning_rate": 0.00029993344075111286,
"loss": 2.0211090087890624,
"step": 940
},
{
"epoch": 0.1216,
"grad_norm": 0.8035023808479309,
"learning_rate": 0.0002999316273684078,
"loss": 2.0083597183227537,
"step": 950
},
{
"epoch": 0.12288,
"grad_norm": 0.8145981431007385,
"learning_rate": 0.0002999297896197363,
"loss": 2.0231433868408204,
"step": 960
},
{
"epoch": 0.12416,
"grad_norm": 0.6948875188827515,
"learning_rate": 0.000299927927505397,
"loss": 2.0229448318481444,
"step": 970
},
{
"epoch": 0.12544,
"grad_norm": 0.7178237438201904,
"learning_rate": 0.00029992604102569256,
"loss": 2.024074363708496,
"step": 980
},
{
"epoch": 0.12672,
"grad_norm": 0.7583261132240295,
"learning_rate": 0.0002999241301809295,
"loss": 2.0232542037963865,
"step": 990
},
{
"epoch": 0.128,
"grad_norm": 0.9775397181510925,
"learning_rate": 0.00029992219497141834,
"loss": 2.0318603515625,
"step": 1000
},
{
"epoch": 0.12928,
"grad_norm": 0.7427451014518738,
"learning_rate": 0.0002999202353974737,
"loss": 2.0150304794311524,
"step": 1010
},
{
"epoch": 0.13056,
"grad_norm": 0.6991038918495178,
"learning_rate": 0.00029991825145941393,
"loss": 2.0047382354736327,
"step": 1020
},
{
"epoch": 0.13184,
"grad_norm": 0.7759991884231567,
"learning_rate": 0.0002999162431575615,
"loss": 2.0170944213867186,
"step": 1030
},
{
"epoch": 0.13312,
"grad_norm": 0.8461285829544067,
"learning_rate": 0.0002999142104922427,
"loss": 2.0140987396240235,
"step": 1040
},
{
"epoch": 0.1344,
"grad_norm": 0.872622549533844,
"learning_rate": 0.000299912153463788,
"loss": 2.0137298583984373,
"step": 1050
},
{
"epoch": 0.13568,
"grad_norm": 0.6421040892601013,
"learning_rate": 0.0002999100720725316,
"loss": 2.0034528732299806,
"step": 1060
},
{
"epoch": 0.13696,
"grad_norm": 0.7797923684120178,
"learning_rate": 0.0002999079663188118,
"loss": 2.0397003173828123,
"step": 1070
},
{
"epoch": 0.13824,
"grad_norm": 0.6871171593666077,
"learning_rate": 0.0002999058362029708,
"loss": 2.0207529067993164,
"step": 1080
},
{
"epoch": 0.13952,
"grad_norm": 0.725261390209198,
"learning_rate": 0.00029990368172535473,
"loss": 2.0161842346191405,
"step": 1090
},
{
"epoch": 0.1408,
"grad_norm": 0.766487717628479,
"learning_rate": 0.0002999015028863138,
"loss": 2.0156953811645506,
"step": 1100
},
{
"epoch": 0.14208,
"grad_norm": 0.6803558468818665,
"learning_rate": 0.000299899299686202,
"loss": 2.029312515258789,
"step": 1110
},
{
"epoch": 0.14336,
"grad_norm": 0.8679984211921692,
"learning_rate": 0.0002998970721253776,
"loss": 2.003293991088867,
"step": 1120
},
{
"epoch": 0.14464,
"grad_norm": 0.8779763579368591,
"learning_rate": 0.0002998948202042023,
"loss": 2.0375221252441404,
"step": 1130
},
{
"epoch": 0.14592,
"grad_norm": 0.7275505065917969,
"learning_rate": 0.00029989254392304225,
"loss": 1.998856544494629,
"step": 1140
},
{
"epoch": 0.1472,
"grad_norm": 0.7902148962020874,
"learning_rate": 0.0002998902432822674,
"loss": 2.0269426345825194,
"step": 1150
},
{
"epoch": 0.14848,
"grad_norm": 0.7712279558181763,
"learning_rate": 0.00029988791828225155,
"loss": 2.010341835021973,
"step": 1160
},
{
"epoch": 0.14976,
"grad_norm": 0.7030239701271057,
"learning_rate": 0.0002998855689233726,
"loss": 2.005113983154297,
"step": 1170
},
{
"epoch": 0.15104,
"grad_norm": 0.6882237195968628,
"learning_rate": 0.00029988319520601235,
"loss": 2.0123071670532227,
"step": 1180
},
{
"epoch": 0.15232,
"grad_norm": 0.711328387260437,
"learning_rate": 0.0002998807971305565,
"loss": 2.019595718383789,
"step": 1190
},
{
"epoch": 0.1536,
"grad_norm": 0.8122184872627258,
"learning_rate": 0.0002998783746973948,
"loss": 2.018142509460449,
"step": 1200
},
{
"epoch": 0.15488,
"grad_norm": 0.8645702004432678,
"learning_rate": 0.000299875927906921,
"loss": 2.0178430557250975,
"step": 1210
},
{
"epoch": 0.15616,
"grad_norm": 0.84963458776474,
"learning_rate": 0.0002998734567595327,
"loss": 2.0299404144287108,
"step": 1220
},
{
"epoch": 0.15744,
"grad_norm": 0.7000954151153564,
"learning_rate": 0.0002998709612556314,
"loss": 2.008827781677246,
"step": 1230
},
{
"epoch": 0.15872,
"grad_norm": 0.7619990110397339,
"learning_rate": 0.0002998684413956228,
"loss": 2.0060449600219727,
"step": 1240
},
{
"epoch": 0.16,
"grad_norm": 0.7732600569725037,
"learning_rate": 0.0002998658971799163,
"loss": 2.0067914962768554,
"step": 1250
},
{
"epoch": 0.16128,
"grad_norm": 0.7497115135192871,
"learning_rate": 0.0002998633286089255,
"loss": 2.006074142456055,
"step": 1260
},
{
"epoch": 0.16256,
"grad_norm": 0.8291988372802734,
"learning_rate": 0.0002998607356830677,
"loss": 2.0243513107299806,
"step": 1270
},
{
"epoch": 0.16384,
"grad_norm": 0.7676268219947815,
"learning_rate": 0.0002998581184027642,
"loss": 2.009158706665039,
"step": 1280
},
{
"epoch": 0.16512,
"grad_norm": 0.6954494714736938,
"learning_rate": 0.0002998554767684406,
"loss": 2.0111303329467773,
"step": 1290
},
{
"epoch": 0.1664,
"grad_norm": 0.8298436999320984,
"learning_rate": 0.0002998528107805261,
"loss": 2.024542236328125,
"step": 1300
},
{
"epoch": 0.16768,
"grad_norm": 0.783103883266449,
"learning_rate": 0.00029985012043945387,
"loss": 2.0101518630981445,
"step": 1310
},
{
"epoch": 0.16896,
"grad_norm": 0.8619032502174377,
"learning_rate": 0.00029984740574566125,
"loss": 2.018630027770996,
"step": 1320
},
{
"epoch": 0.17024,
"grad_norm": 0.7219138145446777,
"learning_rate": 0.0002998446666995893,
"loss": 1.9975326538085938,
"step": 1330
},
{
"epoch": 0.17152,
"grad_norm": 0.7052103281021118,
"learning_rate": 0.00029984190330168325,
"loss": 2.0083797454833983,
"step": 1340
},
{
"epoch": 0.1728,
"grad_norm": 1.1350038051605225,
"learning_rate": 0.00029983911555239213,
"loss": 2.011388397216797,
"step": 1350
},
{
"epoch": 0.17408,
"grad_norm": 0.78801429271698,
"learning_rate": 0.00029983630345216904,
"loss": 2.008464050292969,
"step": 1360
},
{
"epoch": 0.17536,
"grad_norm": 0.7578643560409546,
"learning_rate": 0.0002998334670014709,
"loss": 2.025455284118652,
"step": 1370
},
{
"epoch": 0.17664,
"grad_norm": 0.9882360100746155,
"learning_rate": 0.00029983060620075876,
"loss": 2.0122611999511717,
"step": 1380
},
{
"epoch": 0.17792,
"grad_norm": 0.7348957061767578,
"learning_rate": 0.00029982772105049754,
"loss": 2.018396759033203,
"step": 1390
},
{
"epoch": 0.1792,
"grad_norm": 0.7000976800918579,
"learning_rate": 0.00029982481155115605,
"loss": 2.0135793685913086,
"step": 1400
},
{
"epoch": 0.18048,
"grad_norm": 0.8569319248199463,
"learning_rate": 0.0002998218777032072,
"loss": 2.0004940032958984,
"step": 1410
},
{
"epoch": 0.18176,
"grad_norm": 0.7729265689849854,
"learning_rate": 0.0002998189195071277,
"loss": 1.9858884811401367,
"step": 1420
},
{
"epoch": 0.18304,
"grad_norm": 0.6754656434059143,
"learning_rate": 0.00029981593696339835,
"loss": 2.004491424560547,
"step": 1430
},
{
"epoch": 0.18432,
"grad_norm": 0.7580174803733826,
"learning_rate": 0.0002998129300725038,
"loss": 2.0258563995361327,
"step": 1440
},
{
"epoch": 0.1856,
"grad_norm": 0.8150157928466797,
"learning_rate": 0.0002998098988349328,
"loss": 2.0141185760498046,
"step": 1450
},
{
"epoch": 0.18688,
"grad_norm": 0.8222833275794983,
"learning_rate": 0.00029980684325117793,
"loss": 1.9962928771972657,
"step": 1460
},
{
"epoch": 0.18816,
"grad_norm": 0.7298377156257629,
"learning_rate": 0.00029980376332173575,
"loss": 2.0095237731933593,
"step": 1470
},
{
"epoch": 0.18944,
"grad_norm": 0.7222643494606018,
"learning_rate": 0.00029980065904710673,
"loss": 2.016450119018555,
"step": 1480
},
{
"epoch": 0.19072,
"grad_norm": 0.7348253726959229,
"learning_rate": 0.0002997975304277955,
"loss": 1.9989503860473632,
"step": 1490
},
{
"epoch": 0.192,
"grad_norm": 0.7901546359062195,
"learning_rate": 0.00029979437746431044,
"loss": 2.017726516723633,
"step": 1500
},
{
"epoch": 0.19328,
"grad_norm": 0.7217196822166443,
"learning_rate": 0.0002997912001571639,
"loss": 2.002857208251953,
"step": 1510
},
{
"epoch": 0.19456,
"grad_norm": 0.7372826337814331,
"learning_rate": 0.00029978799850687227,
"loss": 2.018671989440918,
"step": 1520
},
{
"epoch": 0.19584,
"grad_norm": 0.8103359937667847,
"learning_rate": 0.0002997847725139559,
"loss": 2.0113971710205076,
"step": 1530
},
{
"epoch": 0.19712,
"grad_norm": 0.7941066026687622,
"learning_rate": 0.000299781522178939,
"loss": 1.9965641021728515,
"step": 1540
},
{
"epoch": 0.1984,
"grad_norm": 0.7437220215797424,
"learning_rate": 0.00029977824750234983,
"loss": 2.0001338958740233,
"step": 1550
},
{
"epoch": 0.19968,
"grad_norm": 0.8794814348220825,
"learning_rate": 0.0002997749484847206,
"loss": 2.023735237121582,
"step": 1560
},
{
"epoch": 0.20096,
"grad_norm": 0.7125862836837769,
"learning_rate": 0.0002997716251265873,
"loss": 2.0026830673217773,
"step": 1570
},
{
"epoch": 0.20224,
"grad_norm": 0.729263961315155,
"learning_rate": 0.00029976827742849024,
"loss": 2.005397415161133,
"step": 1580
},
{
"epoch": 0.20352,
"grad_norm": 0.7843178510665894,
"learning_rate": 0.00029976490539097335,
"loss": 2.027159881591797,
"step": 1590
},
{
"epoch": 0.2048,
"grad_norm": 0.7313951253890991,
"learning_rate": 0.00029976150901458456,
"loss": 2.0337509155273437,
"step": 1600
},
{
"epoch": 0.20608,
"grad_norm": 0.7452402114868164,
"learning_rate": 0.00029975808829987594,
"loss": 1.9958429336547852,
"step": 1610
},
{
"epoch": 0.20736,
"grad_norm": 0.7813226580619812,
"learning_rate": 0.0002997546432474034,
"loss": 1.9967842102050781,
"step": 1620
},
{
"epoch": 0.20864,
"grad_norm": 0.6847978830337524,
"learning_rate": 0.0002997511738577268,
"loss": 2.0223621368408202,
"step": 1630
},
{
"epoch": 0.20992,
"grad_norm": 0.710649311542511,
"learning_rate": 0.00029974768013140985,
"loss": 2.0229820251464843,
"step": 1640
},
{
"epoch": 0.2112,
"grad_norm": 0.7057698965072632,
"learning_rate": 0.0002997441620690205,
"loss": 2.0189645767211912,
"step": 1650
},
{
"epoch": 0.21248,
"grad_norm": 1.0434540510177612,
"learning_rate": 0.00029974061967113043,
"loss": 2.0161455154418944,
"step": 1660
},
{
"epoch": 0.21376,
"grad_norm": 0.7728395462036133,
"learning_rate": 0.00029973705293831525,
"loss": 2.0083614349365235,
"step": 1670
},
{
"epoch": 0.21504,
"grad_norm": 0.7454256415367126,
"learning_rate": 0.00029973346187115466,
"loss": 2.000332069396973,
"step": 1680
},
{
"epoch": 0.21632,
"grad_norm": 0.6826542019844055,
"learning_rate": 0.0002997298464702323,
"loss": 2.0105188369750975,
"step": 1690
},
{
"epoch": 0.2176,
"grad_norm": 0.8427369594573975,
"learning_rate": 0.00029972620673613564,
"loss": 1.9911138534545898,
"step": 1700
},
{
"epoch": 0.21888,
"grad_norm": 0.8828198313713074,
"learning_rate": 0.00029972254266945623,
"loss": 2.017778778076172,
"step": 1710
},
{
"epoch": 0.22016,
"grad_norm": 0.8785969614982605,
"learning_rate": 0.0002997188542707895,
"loss": 2.0029521942138673,
"step": 1720
},
{
"epoch": 0.22144,
"grad_norm": 0.7189239859580994,
"learning_rate": 0.00029971514154073496,
"loss": 1.9962823867797852,
"step": 1730
},
{
"epoch": 0.22272,
"grad_norm": 0.65921550989151,
"learning_rate": 0.00029971140447989584,
"loss": 2.0048191070556642,
"step": 1740
},
{
"epoch": 0.224,
"grad_norm": 0.8463709950447083,
"learning_rate": 0.0002997076430888796,
"loss": 2.013206672668457,
"step": 1750
},
{
"epoch": 0.22528,
"grad_norm": 0.6320969462394714,
"learning_rate": 0.00029970385736829747,
"loss": 2.0032022476196287,
"step": 1760
},
{
"epoch": 0.22656,
"grad_norm": 0.777824342250824,
"learning_rate": 0.00029970004731876465,
"loss": 1.9946500778198242,
"step": 1770
},
{
"epoch": 0.22784,
"grad_norm": 0.7572101354598999,
"learning_rate": 0.0002996962129409003,
"loss": 2.011198043823242,
"step": 1780
},
{
"epoch": 0.22912,
"grad_norm": 0.6622071266174316,
"learning_rate": 0.00029969235423532765,
"loss": 1.9974157333374023,
"step": 1790
},
{
"epoch": 0.2304,
"grad_norm": 0.6875664591789246,
"learning_rate": 0.0002996884712026737,
"loss": 1.9945417404174806,
"step": 1800
},
{
"epoch": 0.23168,
"grad_norm": 0.7916296124458313,
"learning_rate": 0.0002996845638435696,
"loss": 2.035945510864258,
"step": 1810
},
{
"epoch": 0.23296,
"grad_norm": 0.7541037797927856,
"learning_rate": 0.00029968063215865033,
"loss": 1.9989112854003905,
"step": 1820
},
{
"epoch": 0.23424,
"grad_norm": 0.599982738494873,
"learning_rate": 0.0002996766761485547,
"loss": 1.9939571380615235,
"step": 1830
},
{
"epoch": 0.23552,
"grad_norm": 0.739617645740509,
"learning_rate": 0.00029967269581392575,
"loss": 1.9879533767700195,
"step": 1840
},
{
"epoch": 0.2368,
"grad_norm": 0.7203322649002075,
"learning_rate": 0.00029966869115541036,
"loss": 1.9934770584106445,
"step": 1850
},
{
"epoch": 0.23808,
"grad_norm": 0.9270490407943726,
"learning_rate": 0.00029966466217365927,
"loss": 1.9913738250732422,
"step": 1860
},
{
"epoch": 0.23936,
"grad_norm": 0.8059845566749573,
"learning_rate": 0.0002996606088693273,
"loss": 2.032894325256348,
"step": 1870
},
{
"epoch": 0.24064,
"grad_norm": 0.7291269898414612,
"learning_rate": 0.0002996565312430731,
"loss": 2.0083896636962892,
"step": 1880
},
{
"epoch": 0.24192,
"grad_norm": 0.7227823734283447,
"learning_rate": 0.0002996524292955594,
"loss": 1.998744010925293,
"step": 1890
},
{
"epoch": 0.2432,
"grad_norm": 0.7445799708366394,
"learning_rate": 0.0002996483030274528,
"loss": 1.9974910736083984,
"step": 1900
},
{
"epoch": 0.24448,
"grad_norm": 0.7004072070121765,
"learning_rate": 0.0002996441524394239,
"loss": 2.00911865234375,
"step": 1910
},
{
"epoch": 0.24576,
"grad_norm": 0.9402841329574585,
"learning_rate": 0.00029963997753214717,
"loss": 2.0093734741210936,
"step": 1920
},
{
"epoch": 0.24704,
"grad_norm": 0.6809439659118652,
"learning_rate": 0.0002996357783063011,
"loss": 2.0103471755981444,
"step": 1930
},
{
"epoch": 0.24832,
"grad_norm": 0.7677909731864929,
"learning_rate": 0.00029963155476256826,
"loss": 1.9892677307128905,
"step": 1940
},
{
"epoch": 0.2496,
"grad_norm": 0.8318672180175781,
"learning_rate": 0.0002996273069016348,
"loss": 1.9922115325927734,
"step": 1950
},
{
"epoch": 0.25088,
"grad_norm": 0.6714010834693909,
"learning_rate": 0.00029962303472419124,
"loss": 1.993598747253418,
"step": 1960
},
{
"epoch": 0.25216,
"grad_norm": 0.7420302629470825,
"learning_rate": 0.00029961873823093176,
"loss": 2.010415267944336,
"step": 1970
},
{
"epoch": 0.25344,
"grad_norm": 0.8029727339744568,
"learning_rate": 0.00029961441742255477,
"loss": 2.0113759994506837,
"step": 1980
},
{
"epoch": 0.25472,
"grad_norm": 0.7108703255653381,
"learning_rate": 0.0002996100722997622,
"loss": 2.0091278076171877,
"step": 1990
},
{
"epoch": 0.256,
"grad_norm": 0.6735569834709167,
"learning_rate": 0.0002996057028632604,
"loss": 2.0023717880249023,
"step": 2000
},
{
"epoch": 0.25728,
"grad_norm": 0.8867200016975403,
"learning_rate": 0.0002996013091137594,
"loss": 1.997395896911621,
"step": 2010
},
{
"epoch": 0.25856,
"grad_norm": 0.7353429794311523,
"learning_rate": 0.0002995968910519732,
"loss": 2.015260124206543,
"step": 2020
},
{
"epoch": 0.25984,
"grad_norm": 0.8277012705802917,
"learning_rate": 0.00029959244867861987,
"loss": 1.979180145263672,
"step": 2030
},
{
"epoch": 0.26112,
"grad_norm": 0.6707190871238708,
"learning_rate": 0.00029958798199442136,
"loss": 2.0032121658325197,
"step": 2040
},
{
"epoch": 0.2624,
"grad_norm": 0.734474241733551,
"learning_rate": 0.0002995834910001035,
"loss": 2.0063905715942383,
"step": 2050
},
{
"epoch": 0.26368,
"grad_norm": 0.6844320297241211,
"learning_rate": 0.0002995789756963961,
"loss": 1.9935205459594727,
"step": 2060
},
{
"epoch": 0.26496,
"grad_norm": 0.7805824875831604,
"learning_rate": 0.00029957443608403313,
"loss": 1.9970199584960937,
"step": 2070
},
{
"epoch": 0.26624,
"grad_norm": 0.7733249664306641,
"learning_rate": 0.0002995698721637522,
"loss": 2.0413721084594725,
"step": 2080
},
{
"epoch": 0.26752,
"grad_norm": 1.118902564048767,
"learning_rate": 0.00029956528393629506,
"loss": 2.0041427612304688,
"step": 2090
},
{
"epoch": 0.2688,
"grad_norm": 0.7652356624603271,
"learning_rate": 0.00029956067140240734,
"loss": 2.009052276611328,
"step": 2100
},
{
"epoch": 0.27008,
"grad_norm": 0.7870017290115356,
"learning_rate": 0.0002995560345628387,
"loss": 2.0207101821899416,
"step": 2110
},
{
"epoch": 0.27136,
"grad_norm": 0.5984968543052673,
"learning_rate": 0.0002995513734183426,
"loss": 2.008289337158203,
"step": 2120
},
{
"epoch": 0.27264,
"grad_norm": 0.7700141072273254,
"learning_rate": 0.00029954668796967654,
"loss": 2.0115030288696287,
"step": 2130
},
{
"epoch": 0.27392,
"grad_norm": 1.2647989988327026,
"learning_rate": 0.0002995419782176021,
"loss": 2.023280143737793,
"step": 2140
},
{
"epoch": 0.2752,
"grad_norm": 0.8249433040618896,
"learning_rate": 0.0002995372441628845,
"loss": 1.9807783126831056,
"step": 2150
},
{
"epoch": 0.27648,
"grad_norm": 0.7661439776420593,
"learning_rate": 0.00029953248580629323,
"loss": 2.0067214965820312,
"step": 2160
},
{
"epoch": 0.27776,
"grad_norm": 0.6710767149925232,
"learning_rate": 0.0002995277031486015,
"loss": 2.005890464782715,
"step": 2170
},
{
"epoch": 0.27904,
"grad_norm": 0.8246187567710876,
"learning_rate": 0.00029952289619058673,
"loss": 2.016609764099121,
"step": 2180
},
{
"epoch": 0.28032,
"grad_norm": 0.6944814324378967,
"learning_rate": 0.0002995180649330299,
"loss": 2.0185358047485353,
"step": 2190
},
{
"epoch": 0.2816,
"grad_norm": 1.15223228931427,
"learning_rate": 0.0002995132093767162,
"loss": 1.9930595397949218,
"step": 2200
},
{
"epoch": 0.28288,
"grad_norm": 1.1536227464675903,
"learning_rate": 0.00029950832952243487,
"loss": 2.0209888458251952,
"step": 2210
},
{
"epoch": 0.28416,
"grad_norm": 0.8789283037185669,
"learning_rate": 0.00029950342537097874,
"loss": 2.014572334289551,
"step": 2220
},
{
"epoch": 0.28544,
"grad_norm": 0.8173903226852417,
"learning_rate": 0.00029949849692314503,
"loss": 2.0143598556518554,
"step": 2230
},
{
"epoch": 0.28672,
"grad_norm": 0.7095955014228821,
"learning_rate": 0.0002994935441797345,
"loss": 2.004468536376953,
"step": 2240
},
{
"epoch": 0.288,
"grad_norm": 0.8468304872512817,
"learning_rate": 0.0002994885671415521,
"loss": 2.037894630432129,
"step": 2250
},
{
"epoch": 0.28928,
"grad_norm": 0.8071026802062988,
"learning_rate": 0.0002994835658094067,
"loss": 2.023789405822754,
"step": 2260
},
{
"epoch": 0.29056,
"grad_norm": 0.7277827262878418,
"learning_rate": 0.0002994785401841111,
"loss": 2.010818672180176,
"step": 2270
},
{
"epoch": 0.29184,
"grad_norm": 0.7613065838813782,
"learning_rate": 0.00029947349026648184,
"loss": 1.9951431274414062,
"step": 2280
},
{
"epoch": 0.29312,
"grad_norm": 0.7575923800468445,
"learning_rate": 0.0002994684160573399,
"loss": 2.0147769927978514,
"step": 2290
},
{
"epoch": 0.2944,
"grad_norm": 1.050876259803772,
"learning_rate": 0.0002994633175575097,
"loss": 2.0217056274414062,
"step": 2300
},
{
"epoch": 0.29568,
"grad_norm": 0.6943807601928711,
"learning_rate": 0.0002994581947678199,
"loss": 2.0191987991333007,
"step": 2310
},
{
"epoch": 0.29696,
"grad_norm": 0.8438522219657898,
"learning_rate": 0.00029945304768910297,
"loss": 1.9989118576049805,
"step": 2320
},
{
"epoch": 0.29824,
"grad_norm": 0.7962123155593872,
"learning_rate": 0.00029944787632219547,
"loss": 2.0173967361450194,
"step": 2330
},
{
"epoch": 0.29952,
"grad_norm": 0.7668672204017639,
"learning_rate": 0.0002994426806679378,
"loss": 2.0132951736450195,
"step": 2340
},
{
"epoch": 0.3008,
"grad_norm": 0.6691519618034363,
"learning_rate": 0.0002994374607271742,
"loss": 2.0168371200561523,
"step": 2350
},
{
"epoch": 0.30208,
"grad_norm": 0.6511347889900208,
"learning_rate": 0.0002994322165007531,
"loss": 2.0063940048217774,
"step": 2360
},
{
"epoch": 0.30336,
"grad_norm": 0.8013675212860107,
"learning_rate": 0.00029942694798952677,
"loss": 2.0087839126586915,
"step": 2370
},
{
"epoch": 0.30464,
"grad_norm": 0.899410605430603,
"learning_rate": 0.0002994216551943513,
"loss": 2.0143436431884765,
"step": 2380
},
{
"epoch": 0.30592,
"grad_norm": 0.6686886548995972,
"learning_rate": 0.000299416338116087,
"loss": 1.9976930618286133,
"step": 2390
},
{
"epoch": 0.3072,
"grad_norm": 0.8113393187522888,
"learning_rate": 0.0002994109967555978,
"loss": 2.0220170974731446,
"step": 2400
},
{
"epoch": 0.30848,
"grad_norm": 0.7538885474205017,
"learning_rate": 0.00029940563111375195,
"loss": 2.004685974121094,
"step": 2410
},
{
"epoch": 0.30976,
"grad_norm": 0.8250473141670227,
"learning_rate": 0.0002994002411914213,
"loss": 2.0201698303222657,
"step": 2420
},
{
"epoch": 0.31104,
"grad_norm": 0.7206689119338989,
"learning_rate": 0.00029939482698948175,
"loss": 1.9911544799804688,
"step": 2430
},
{
"epoch": 0.31232,
"grad_norm": 0.7519621253013611,
"learning_rate": 0.0002993893885088133,
"loss": 2.0177082061767577,
"step": 2440
},
{
"epoch": 0.3136,
"grad_norm": 0.777955174446106,
"learning_rate": 0.0002993839257502997,
"loss": 1.9929460525512694,
"step": 2450
},
{
"epoch": 0.31488,
"grad_norm": 0.7577837705612183,
"learning_rate": 0.0002993784387148288,
"loss": 2.0109020233154298,
"step": 2460
},
{
"epoch": 0.31616,
"grad_norm": 0.7473259568214417,
"learning_rate": 0.0002993729274032922,
"loss": 2.0051565170288086,
"step": 2470
},
{
"epoch": 0.31744,
"grad_norm": 0.7150516510009766,
"learning_rate": 0.0002993673918165856,
"loss": 1.988702392578125,
"step": 2480
},
{
"epoch": 0.31872,
"grad_norm": 0.7589870095252991,
"learning_rate": 0.00029936183195560873,
"loss": 2.0132009506225588,
"step": 2490
},
{
"epoch": 0.32,
"grad_norm": 0.8448571562767029,
"learning_rate": 0.00029935624782126505,
"loss": 2.00791072845459,
"step": 2500
},
{
"epoch": 0.32128,
"grad_norm": 0.9435201287269592,
"learning_rate": 0.000299350639414462,
"loss": 2.0194942474365236,
"step": 2510
},
{
"epoch": 0.32256,
"grad_norm": 0.65596604347229,
"learning_rate": 0.00029934500673611114,
"loss": 2.0138187408447266,
"step": 2520
},
{
"epoch": 0.32384,
"grad_norm": 0.7170972228050232,
"learning_rate": 0.00029933934978712784,
"loss": 2.015252876281738,
"step": 2530
},
{
"epoch": 0.32512,
"grad_norm": 0.7384889125823975,
"learning_rate": 0.00029933366856843135,
"loss": 2.0059532165527343,
"step": 2540
},
{
"epoch": 0.3264,
"grad_norm": 0.7757416367530823,
"learning_rate": 0.000299327963080945,
"loss": 2.024062728881836,
"step": 2550
},
{
"epoch": 0.32768,
"grad_norm": 0.8023697733879089,
"learning_rate": 0.000299322233325596,
"loss": 1.9900062561035157,
"step": 2560
},
{
"epoch": 0.32896,
"grad_norm": 0.7611852288246155,
"learning_rate": 0.00029931647930331556,
"loss": 2.0226030349731445,
"step": 2570
},
{
"epoch": 0.33024,
"grad_norm": 0.8219276070594788,
"learning_rate": 0.0002993107010150387,
"loss": 1.9970281600952149,
"step": 2580
},
{
"epoch": 0.33152,
"grad_norm": 0.7846634984016418,
"learning_rate": 0.0002993048984617046,
"loss": 1.9928812026977538,
"step": 2590
},
{
"epoch": 0.3328,
"grad_norm": 0.7521805167198181,
"learning_rate": 0.0002992990716442561,
"loss": 1.99877986907959,
"step": 2600
},
{
"epoch": 0.33408,
"grad_norm": 0.660119354724884,
"learning_rate": 0.0002992932205636403,
"loss": 2.007124900817871,
"step": 2610
},
{
"epoch": 0.33536,
"grad_norm": 0.7946996688842773,
"learning_rate": 0.0002992873452208079,
"loss": 1.9985849380493164,
"step": 2620
},
{
"epoch": 0.33664,
"grad_norm": 0.8377641439437866,
"learning_rate": 0.00029928144561671387,
"loss": 2.0259756088256835,
"step": 2630
},
{
"epoch": 0.33792,
"grad_norm": 0.6834446787834167,
"learning_rate": 0.0002992755217523169,
"loss": 1.9996057510375977,
"step": 2640
},
{
"epoch": 0.3392,
"grad_norm": 0.8117668628692627,
"learning_rate": 0.00029926957362857976,
"loss": 1.9947677612304688,
"step": 2650
},
{
"epoch": 0.34048,
"grad_norm": 0.6758447289466858,
"learning_rate": 0.0002992636012464691,
"loss": 1.984456443786621,
"step": 2660
},
{
"epoch": 0.34176,
"grad_norm": 0.6741154789924622,
"learning_rate": 0.00029925760460695547,
"loss": 1.9969703674316406,
"step": 2670
},
{
"epoch": 0.34304,
"grad_norm": 0.7465670108795166,
"learning_rate": 0.0002992515837110134,
"loss": 2.0042219161987305,
"step": 2680
},
{
"epoch": 0.34432,
"grad_norm": 0.8404413461685181,
"learning_rate": 0.0002992455385596214,
"loss": 1.9949943542480468,
"step": 2690
},
{
"epoch": 0.3456,
"grad_norm": 0.7381366491317749,
"learning_rate": 0.00029923946915376187,
"loss": 2.0078285217285154,
"step": 2700
},
{
"epoch": 0.34688,
"grad_norm": 0.7588344216346741,
"learning_rate": 0.00029923337549442124,
"loss": 2.008309173583984,
"step": 2710
},
{
"epoch": 0.34816,
"grad_norm": 0.7521999478340149,
"learning_rate": 0.0002992272575825897,
"loss": 2.0139354705810546,
"step": 2720
},
{
"epoch": 0.34944,
"grad_norm": 0.6759751439094543,
"learning_rate": 0.0002992211154192616,
"loss": 2.001015853881836,
"step": 2730
},
{
"epoch": 0.35072,
"grad_norm": 0.6119874119758606,
"learning_rate": 0.00029921494900543507,
"loss": 2.0114923477172852,
"step": 2740
},
{
"epoch": 0.352,
"grad_norm": 0.7517712116241455,
"learning_rate": 0.0002992087583421123,
"loss": 1.9976957321166993,
"step": 2750
},
{
"epoch": 0.35328,
"grad_norm": 0.769814133644104,
"learning_rate": 0.0002992025434302992,
"loss": 1.9972566604614257,
"step": 2760
},
{
"epoch": 0.35456,
"grad_norm": 0.6553206443786621,
"learning_rate": 0.00029919630427100596,
"loss": 2.007891845703125,
"step": 2770
},
{
"epoch": 0.35584,
"grad_norm": 0.6894077658653259,
"learning_rate": 0.0002991900408652465,
"loss": 2.0165834426879883,
"step": 2780
},
{
"epoch": 0.35712,
"grad_norm": 0.669863224029541,
"learning_rate": 0.00029918375321403857,
"loss": 2.010799789428711,
"step": 2790
},
{
"epoch": 0.3584,
"grad_norm": 0.7974678874015808,
"learning_rate": 0.00029917744131840416,
"loss": 2.0101280212402344,
"step": 2800
},
{
"epoch": 0.35968,
"grad_norm": 0.7307773232460022,
"learning_rate": 0.00029917110517936894,
"loss": 2.024929428100586,
"step": 2810
},
{
"epoch": 0.36096,
"grad_norm": 0.7141740322113037,
"learning_rate": 0.00029916474479796265,
"loss": 2.0083641052246093,
"step": 2820
},
{
"epoch": 0.36224,
"grad_norm": 0.6874279379844666,
"learning_rate": 0.000299158360175219,
"loss": 2.0239688873291017,
"step": 2830
},
{
"epoch": 0.36352,
"grad_norm": 0.6707704663276672,
"learning_rate": 0.0002991519513121755,
"loss": 2.005034828186035,
"step": 2840
},
{
"epoch": 0.3648,
"grad_norm": 0.8018107414245605,
"learning_rate": 0.00029914551820987373,
"loss": 1.997845458984375,
"step": 2850
},
{
"epoch": 0.36608,
"grad_norm": 0.7515741586685181,
"learning_rate": 0.00029913906086935916,
"loss": 2.0154483795166014,
"step": 2860
},
{
"epoch": 0.36736,
"grad_norm": 0.7631517648696899,
"learning_rate": 0.0002991325792916811,
"loss": 1.977918815612793,
"step": 2870
},
{
"epoch": 0.36864,
"grad_norm": 0.8842434287071228,
"learning_rate": 0.000299126073477893,
"loss": 1.975722885131836,
"step": 2880
},
{
"epoch": 0.36992,
"grad_norm": 0.7274171113967896,
"learning_rate": 0.00029911954342905217,
"loss": 1.9984859466552733,
"step": 2890
},
{
"epoch": 0.3712,
"grad_norm": 0.7220802307128906,
"learning_rate": 0.00029911298914621977,
"loss": 2.0087894439697265,
"step": 2900
},
{
"epoch": 0.37248,
"grad_norm": 0.6422837376594543,
"learning_rate": 0.0002991064106304609,
"loss": 2.002351760864258,
"step": 2910
},
{
"epoch": 0.37376,
"grad_norm": 0.753294825553894,
"learning_rate": 0.00029909980788284476,
"loss": 1.986056137084961,
"step": 2920
},
{
"epoch": 0.37504,
"grad_norm": 0.7144089937210083,
"learning_rate": 0.00029909318090444437,
"loss": 1.9991958618164063,
"step": 2930
},
{
"epoch": 0.37632,
"grad_norm": 0.704556405544281,
"learning_rate": 0.0002990865296963367,
"loss": 1.9974645614624023,
"step": 2940
},
{
"epoch": 0.3776,
"grad_norm": 0.7290600538253784,
"learning_rate": 0.0002990798542596026,
"loss": 2.015464019775391,
"step": 2950
},
{
"epoch": 0.37888,
"grad_norm": 0.7425021529197693,
"learning_rate": 0.00029907315459532706,
"loss": 2.0130130767822267,
"step": 2960
},
{
"epoch": 0.38016,
"grad_norm": 0.7028226852416992,
"learning_rate": 0.0002990664307045988,
"loss": 1.999576950073242,
"step": 2970
},
{
"epoch": 0.38144,
"grad_norm": 0.6917806267738342,
"learning_rate": 0.00029905968258851047,
"loss": 2.00009765625,
"step": 2980
},
{
"epoch": 0.38272,
"grad_norm": 0.7476458549499512,
"learning_rate": 0.00029905291024815876,
"loss": 1.9876760482788085,
"step": 2990
},
{
"epoch": 0.384,
"grad_norm": 0.8785274028778076,
"learning_rate": 0.00029904611368464435,
"loss": 2.018217086791992,
"step": 3000
},
{
"epoch": 0.38528,
"grad_norm": 0.7297069430351257,
"learning_rate": 0.0002990392928990717,
"loss": 1.9931718826293945,
"step": 3010
},
{
"epoch": 0.38656,
"grad_norm": 0.6672376990318298,
"learning_rate": 0.0002990324478925494,
"loss": 2.0107067108154295,
"step": 3020
},
{
"epoch": 0.38784,
"grad_norm": 0.7302089929580688,
"learning_rate": 0.0002990255786661897,
"loss": 1.994314956665039,
"step": 3030
},
{
"epoch": 0.38912,
"grad_norm": 0.7202204465866089,
"learning_rate": 0.00029901868522110904,
"loss": 2.0174509048461915,
"step": 3040
},
{
"epoch": 0.3904,
"grad_norm": 0.8425076603889465,
"learning_rate": 0.00029901176755842764,
"loss": 2.0013553619384767,
"step": 3050
},
{
"epoch": 0.39168,
"grad_norm": 0.8686426877975464,
"learning_rate": 0.00029900482567926976,
"loss": 2.007709503173828,
"step": 3060
},
{
"epoch": 0.39296,
"grad_norm": 0.6840008497238159,
"learning_rate": 0.00029899785958476357,
"loss": 2.004677963256836,
"step": 3070
},
{
"epoch": 0.39424,
"grad_norm": 0.6260886192321777,
"learning_rate": 0.0002989908692760411,
"loss": 1.9879840850830077,
"step": 3080
},
{
"epoch": 0.39552,
"grad_norm": 0.6933361887931824,
"learning_rate": 0.0002989838547542384,
"loss": 2.011644744873047,
"step": 3090
},
{
"epoch": 0.3968,
"grad_norm": 0.7048612833023071,
"learning_rate": 0.0002989768160204955,
"loss": 2.019415092468262,
"step": 3100
},
{
"epoch": 0.39808,
"grad_norm": 0.7651655673980713,
"learning_rate": 0.00029896975307595614,
"loss": 2.0010753631591798,
"step": 3110
},
{
"epoch": 0.39936,
"grad_norm": 0.6941404938697815,
"learning_rate": 0.00029896266592176825,
"loss": 1.984878921508789,
"step": 3120
},
{
"epoch": 0.40064,
"grad_norm": 0.8553006649017334,
"learning_rate": 0.00029895555455908355,
"loss": 1.9996231079101563,
"step": 3130
},
{
"epoch": 0.40192,
"grad_norm": 0.7001233100891113,
"learning_rate": 0.00029894841898905775,
"loss": 1.9921398162841797,
"step": 3140
},
{
"epoch": 0.4032,
"grad_norm": 0.8051826357841492,
"learning_rate": 0.0002989412592128505,
"loss": 1.999976921081543,
"step": 3150
},
{
"epoch": 0.40448,
"grad_norm": 0.7095255255699158,
"learning_rate": 0.00029893407523162534,
"loss": 2.0126005172729493,
"step": 3160
},
{
"epoch": 0.40576,
"grad_norm": 0.9202665686607361,
"learning_rate": 0.00029892686704654977,
"loss": 1.9911052703857421,
"step": 3170
},
{
"epoch": 0.40704,
"grad_norm": 0.7405928373336792,
"learning_rate": 0.0002989196346587952,
"loss": 1.9965196609497071,
"step": 3180
},
{
"epoch": 0.40832,
"grad_norm": 1.4823166131973267,
"learning_rate": 0.0002989123780695371,
"loss": 2.0123992919921876,
"step": 3190
},
{
"epoch": 0.4096,
"grad_norm": 0.7466282248497009,
"learning_rate": 0.00029890509727995457,
"loss": 2.0086360931396485,
"step": 3200
},
{
"epoch": 0.41088,
"grad_norm": 0.7171423435211182,
"learning_rate": 0.00029889779229123095,
"loss": 2.0146774291992187,
"step": 3210
},
{
"epoch": 0.41216,
"grad_norm": 0.8924009203910828,
"learning_rate": 0.00029889046310455346,
"loss": 2.006013107299805,
"step": 3220
},
{
"epoch": 0.41344,
"grad_norm": 0.6809982657432556,
"learning_rate": 0.0002988831097211131,
"loss": 2.007125473022461,
"step": 3230
},
{
"epoch": 0.41472,
"grad_norm": 0.6054995656013489,
"learning_rate": 0.000298875732142105,
"loss": 1.9968708038330079,
"step": 3240
},
{
"epoch": 0.416,
"grad_norm": 0.6415665149688721,
"learning_rate": 0.000298868330368728,
"loss": 2.004538345336914,
"step": 3250
},
{
"epoch": 0.41728,
"grad_norm": 0.738610029220581,
"learning_rate": 0.00029886090440218505,
"loss": 1.9990394592285157,
"step": 3260
},
{
"epoch": 0.41856,
"grad_norm": 0.7058972120285034,
"learning_rate": 0.00029885345424368294,
"loss": 2.0169488906860353,
"step": 3270
},
{
"epoch": 0.41984,
"grad_norm": 0.6690015196800232,
"learning_rate": 0.0002988459798944324,
"loss": 1.990290069580078,
"step": 3280
},
{
"epoch": 0.42112,
"grad_norm": 0.627063512802124,
"learning_rate": 0.00029883848135564827,
"loss": 1.992024040222168,
"step": 3290
},
{
"epoch": 0.4224,
"grad_norm": 0.6873062252998352,
"learning_rate": 0.000298830958628549,
"loss": 2.018503189086914,
"step": 3300
},
{
"epoch": 0.42368,
"grad_norm": 0.7181523442268372,
"learning_rate": 0.0002988234117143572,
"loss": 2.0042495727539062,
"step": 3310
},
{
"epoch": 0.42496,
"grad_norm": 0.7824463248252869,
"learning_rate": 0.0002988158406142994,
"loss": 1.9997383117675782,
"step": 3320
},
{
"epoch": 0.42624,
"grad_norm": 0.7802339196205139,
"learning_rate": 0.0002988082453296059,
"loss": 1.996378517150879,
"step": 3330
},
{
"epoch": 0.42752,
"grad_norm": 0.7132096886634827,
"learning_rate": 0.00029880062586151105,
"loss": 2.0132312774658203,
"step": 3340
},
{
"epoch": 0.4288,
"grad_norm": 0.8338831067085266,
"learning_rate": 0.00029879298221125323,
"loss": 2.010465621948242,
"step": 3350
},
{
"epoch": 0.43008,
"grad_norm": 0.7584986090660095,
"learning_rate": 0.0002987853143800745,
"loss": 1.9958574295043945,
"step": 3360
},
{
"epoch": 0.43136,
"grad_norm": 0.6547184586524963,
"learning_rate": 0.00029877762236922117,
"loss": 1.9954259872436524,
"step": 3370
},
{
"epoch": 0.43264,
"grad_norm": 0.7016308903694153,
"learning_rate": 0.0002987699061799431,
"loss": 1.9987804412841796,
"step": 3380
},
{
"epoch": 0.43392,
"grad_norm": 0.7248575687408447,
"learning_rate": 0.0002987621658134944,
"loss": 2.0054561614990236,
"step": 3390
},
{
"epoch": 0.4352,
"grad_norm": 1.0197134017944336,
"learning_rate": 0.00029875440127113295,
"loss": 2.0060552597045898,
"step": 3400
},
{
"epoch": 0.43648,
"grad_norm": 0.7604638338088989,
"learning_rate": 0.00029874661255412056,
"loss": 1.9959848403930665,
"step": 3410
},
{
"epoch": 0.43776,
"grad_norm": 0.7341379523277283,
"learning_rate": 0.0002987387996637231,
"loss": 2.0023780822753907,
"step": 3420
},
{
"epoch": 0.43904,
"grad_norm": 0.6850576400756836,
"learning_rate": 0.0002987309626012102,
"loss": 2.020222282409668,
"step": 3430
},
{
"epoch": 0.44032,
"grad_norm": 0.7283312082290649,
"learning_rate": 0.00029872310136785544,
"loss": 1.994460105895996,
"step": 3440
},
{
"epoch": 0.4416,
"grad_norm": 0.7015148997306824,
"learning_rate": 0.00029871521596493654,
"loss": 2.015683174133301,
"step": 3450
},
{
"epoch": 0.44288,
"grad_norm": 0.6451849341392517,
"learning_rate": 0.0002987073063937348,
"loss": 1.9941566467285157,
"step": 3460
},
{
"epoch": 0.44416,
"grad_norm": 0.6728188991546631,
"learning_rate": 0.0002986993726555358,
"loss": 2.0009782791137694,
"step": 3470
},
{
"epoch": 0.44544,
"grad_norm": 0.7595881223678589,
"learning_rate": 0.0002986914147516288,
"loss": 1.994197463989258,
"step": 3480
},
{
"epoch": 0.44672,
"grad_norm": 0.7372469902038574,
"learning_rate": 0.00029868343268330704,
"loss": 1.9918861389160156,
"step": 3490
},
{
"epoch": 0.448,
"grad_norm": 0.8467145562171936,
"learning_rate": 0.0002986754264518678,
"loss": 1.970982551574707,
"step": 3500
},
{
"epoch": 0.44928,
"grad_norm": 0.6036944389343262,
"learning_rate": 0.00029866739605861214,
"loss": 1.970067024230957,
"step": 3510
},
{
"epoch": 0.45056,
"grad_norm": 0.7512437701225281,
"learning_rate": 0.00029865934150484513,
"loss": 1.9916757583618163,
"step": 3520
},
{
"epoch": 0.45184,
"grad_norm": 0.7820079326629639,
"learning_rate": 0.00029865126279187575,
"loss": 2.015567970275879,
"step": 3530
},
{
"epoch": 0.45312,
"grad_norm": 0.753780722618103,
"learning_rate": 0.0002986431599210169,
"loss": 1.9925758361816406,
"step": 3540
},
{
"epoch": 0.4544,
"grad_norm": 0.583357036113739,
"learning_rate": 0.00029863503289358543,
"loss": 2.0011922836303713,
"step": 3550
},
{
"epoch": 0.45568,
"grad_norm": 0.7955908179283142,
"learning_rate": 0.00029862688171090207,
"loss": 1.9962081909179688,
"step": 3560
},
{
"epoch": 0.45696,
"grad_norm": 0.6503195762634277,
"learning_rate": 0.0002986187063742914,
"loss": 2.0152355194091798,
"step": 3570
},
{
"epoch": 0.45824,
"grad_norm": 0.702178955078125,
"learning_rate": 0.00029861050688508224,
"loss": 2.010386848449707,
"step": 3580
},
{
"epoch": 0.45952,
"grad_norm": 0.6806907057762146,
"learning_rate": 0.000298602283244607,
"loss": 2.002973747253418,
"step": 3590
},
{
"epoch": 0.4608,
"grad_norm": 0.6157857775688171,
"learning_rate": 0.00029859403545420214,
"loss": 1.995046615600586,
"step": 3600
},
{
"epoch": 0.46208,
"grad_norm": 0.6844906806945801,
"learning_rate": 0.000298585763515208,
"loss": 2.016707992553711,
"step": 3610
},
{
"epoch": 0.46336,
"grad_norm": 0.7760967016220093,
"learning_rate": 0.00029857746742896897,
"loss": 1.9860107421875,
"step": 3620
},
{
"epoch": 0.46464,
"grad_norm": 0.8712556958198547,
"learning_rate": 0.0002985691471968332,
"loss": 2.001064491271973,
"step": 3630
},
{
"epoch": 0.46592,
"grad_norm": 0.7285525798797607,
"learning_rate": 0.00029856080282015293,
"loss": 2.0140453338623048,
"step": 3640
},
{
"epoch": 0.4672,
"grad_norm": 0.7722259759902954,
"learning_rate": 0.00029855243430028415,
"loss": 2.0120145797729494,
"step": 3650
},
{
"epoch": 0.46848,
"grad_norm": 0.7177460193634033,
"learning_rate": 0.0002985440416385869,
"loss": 1.9976810455322265,
"step": 3660
},
{
"epoch": 0.46976,
"grad_norm": 0.8264991044998169,
"learning_rate": 0.0002985356248364251,
"loss": 2.005207633972168,
"step": 3670
},
{
"epoch": 0.47104,
"grad_norm": 0.8237878084182739,
"learning_rate": 0.00029852718389516666,
"loss": 1.988731575012207,
"step": 3680
},
{
"epoch": 0.47232,
"grad_norm": 1.3742607831954956,
"learning_rate": 0.00029851871881618323,
"loss": 2.0098743438720703,
"step": 3690
},
{
"epoch": 0.4736,
"grad_norm": 0.7714928388595581,
"learning_rate": 0.0002985102296008505,
"loss": 1.9839420318603516,
"step": 3700
},
{
"epoch": 0.47488,
"grad_norm": 0.7396987080574036,
"learning_rate": 0.0002985017162505483,
"loss": 1.9785545349121094,
"step": 3710
},
{
"epoch": 0.47616,
"grad_norm": 0.7527230978012085,
"learning_rate": 0.0002984931787666599,
"loss": 2.0031604766845703,
"step": 3720
},
{
"epoch": 0.47744,
"grad_norm": 0.7075295448303223,
"learning_rate": 0.0002984846171505729,
"loss": 1.9907520294189454,
"step": 3730
},
{
"epoch": 0.47872,
"grad_norm": 0.6408486366271973,
"learning_rate": 0.0002984760314036787,
"loss": 2.024155044555664,
"step": 3740
},
{
"epoch": 0.48,
"grad_norm": 0.7029430866241455,
"learning_rate": 0.0002984674215273725,
"loss": 2.00035400390625,
"step": 3750
},
{
"epoch": 0.48128,
"grad_norm": 0.7707000970840454,
"learning_rate": 0.0002984587875230536,
"loss": 2.0184585571289064,
"step": 3760
},
{
"epoch": 0.48256,
"grad_norm": 0.6695795655250549,
"learning_rate": 0.00029845012939212507,
"loss": 1.9930442810058593,
"step": 3770
},
{
"epoch": 0.48384,
"grad_norm": 0.6928637623786926,
"learning_rate": 0.0002984414471359941,
"loss": 1.9891483306884765,
"step": 3780
},
{
"epoch": 0.48512,
"grad_norm": 0.7137778997421265,
"learning_rate": 0.00029843274075607167,
"loss": 2.0126466751098633,
"step": 3790
},
{
"epoch": 0.4864,
"grad_norm": 0.7941726446151733,
"learning_rate": 0.00029842401025377254,
"loss": 1.9977178573608398,
"step": 3800
},
{
"epoch": 0.48768,
"grad_norm": 0.7112129926681519,
"learning_rate": 0.0002984152556305156,
"loss": 2.0004758834838867,
"step": 3810
},
{
"epoch": 0.48896,
"grad_norm": 0.6955267190933228,
"learning_rate": 0.0002984064768877237,
"loss": 2.0171720504760744,
"step": 3820
},
{
"epoch": 0.49024,
"grad_norm": 0.7090415358543396,
"learning_rate": 0.00029839767402682334,
"loss": 2.005842399597168,
"step": 3830
},
{
"epoch": 0.49152,
"grad_norm": 0.8107616305351257,
"learning_rate": 0.00029838884704924527,
"loss": 2.0197629928588867,
"step": 3840
},
{
"epoch": 0.4928,
"grad_norm": 0.7303240299224854,
"learning_rate": 0.0002983799959564239,
"loss": 2.0161752700805664,
"step": 3850
},
{
"epoch": 0.49408,
"grad_norm": 0.7189793586730957,
"learning_rate": 0.0002983711207497977,
"loss": 2.0196367263793946,
"step": 3860
},
{
"epoch": 0.49536,
"grad_norm": 0.7447757124900818,
"learning_rate": 0.000298362221430809,
"loss": 1.9797452926635741,
"step": 3870
},
{
"epoch": 0.49664,
"grad_norm": 0.7726782560348511,
"learning_rate": 0.000298353298000904,
"loss": 1.994917869567871,
"step": 3880
},
{
"epoch": 0.49792,
"grad_norm": 0.7866300344467163,
"learning_rate": 0.000298344350461533,
"loss": 1.994272232055664,
"step": 3890
},
{
"epoch": 0.4992,
"grad_norm": 0.7009814381599426,
"learning_rate": 0.00029833537881414997,
"loss": 1.9947071075439453,
"step": 3900
},
{
"epoch": 0.50048,
"grad_norm": 0.786959171295166,
"learning_rate": 0.000298326383060213,
"loss": 1.9846187591552735,
"step": 3910
},
{
"epoch": 0.50176,
"grad_norm": 0.798842191696167,
"learning_rate": 0.000298317363201184,
"loss": 2.0009033203125,
"step": 3920
},
{
"epoch": 0.50304,
"grad_norm": 0.6344483494758606,
"learning_rate": 0.0002983083192385289,
"loss": 2.0058782577514647,
"step": 3930
},
{
"epoch": 0.50432,
"grad_norm": 0.7020363211631775,
"learning_rate": 0.0002982992511737174,
"loss": 2.0104549407958983,
"step": 3940
},
{
"epoch": 0.5056,
"grad_norm": 0.7292479872703552,
"learning_rate": 0.00029829015900822314,
"loss": 2.0061105728149413,
"step": 3950
},
{
"epoch": 0.50688,
"grad_norm": 0.7374621629714966,
"learning_rate": 0.00029828104274352385,
"loss": 1.991245651245117,
"step": 3960
},
{
"epoch": 0.50816,
"grad_norm": 0.7191388010978699,
"learning_rate": 0.0002982719023811009,
"loss": 2.005914878845215,
"step": 3970
},
{
"epoch": 0.50944,
"grad_norm": 0.7209792137145996,
"learning_rate": 0.0002982627379224399,
"loss": 1.9991714477539062,
"step": 3980
},
{
"epoch": 0.51072,
"grad_norm": 0.7501109838485718,
"learning_rate": 0.00029825354936903,
"loss": 2.0138620376586913,
"step": 3990
},
{
"epoch": 0.512,
"grad_norm": 0.7204382419586182,
"learning_rate": 0.0002982443367223646,
"loss": 2.002537155151367,
"step": 4000
},
{
"epoch": 0.51328,
"grad_norm": 0.6654132008552551,
"learning_rate": 0.0002982350999839409,
"loss": 2.0058666229248048,
"step": 4010
},
{
"epoch": 0.51456,
"grad_norm": 0.6496480107307434,
"learning_rate": 0.00029822583915525987,
"loss": 2.0121301651000976,
"step": 4020
},
{
"epoch": 0.51584,
"grad_norm": 0.7528657913208008,
"learning_rate": 0.00029821655423782666,
"loss": 1.997304916381836,
"step": 4030
},
{
"epoch": 0.51712,
"grad_norm": 0.7558349967002869,
"learning_rate": 0.0002982072452331501,
"loss": 2.0115285873413087,
"step": 4040
},
{
"epoch": 0.5184,
"grad_norm": 0.649427056312561,
"learning_rate": 0.0002981979121427431,
"loss": 1.9857589721679687,
"step": 4050
},
{
"epoch": 0.51968,
"grad_norm": 0.8884936571121216,
"learning_rate": 0.0002981885549681224,
"loss": 1.9965480804443358,
"step": 4060
},
{
"epoch": 0.52096,
"grad_norm": 0.6860265135765076,
"learning_rate": 0.00029817917371080863,
"loss": 1.9892051696777344,
"step": 4070
},
{
"epoch": 0.52224,
"grad_norm": 0.6648100018501282,
"learning_rate": 0.00029816976837232647,
"loss": 1.9910903930664063,
"step": 4080
},
{
"epoch": 0.52352,
"grad_norm": 0.669529139995575,
"learning_rate": 0.0002981603389542043,
"loss": 1.9876266479492188,
"step": 4090
},
{
"epoch": 0.5248,
"grad_norm": 0.7481969594955444,
"learning_rate": 0.0002981508854579746,
"loss": 2.0230886459350588,
"step": 4100
},
{
"epoch": 0.52608,
"grad_norm": 0.7897581458091736,
"learning_rate": 0.00029814140788517364,
"loss": 1.9916162490844727,
"step": 4110
},
{
"epoch": 0.52736,
"grad_norm": 0.6596736907958984,
"learning_rate": 0.0002981319062373417,
"loss": 1.9848915100097657,
"step": 4120
},
{
"epoch": 0.52864,
"grad_norm": 0.6131610870361328,
"learning_rate": 0.00029812238051602294,
"loss": 2.003184700012207,
"step": 4130
},
{
"epoch": 0.52992,
"grad_norm": 0.775517463684082,
"learning_rate": 0.00029811283072276543,
"loss": 2.0039646148681642,
"step": 4140
},
{
"epoch": 0.5312,
"grad_norm": 0.6710156798362732,
"learning_rate": 0.00029810325685912113,
"loss": 2.0029441833496096,
"step": 4150
},
{
"epoch": 0.53248,
"grad_norm": 0.6885665655136108,
"learning_rate": 0.00029809365892664587,
"loss": 1.9940208435058593,
"step": 4160
},
{
"epoch": 0.53376,
"grad_norm": 0.7027019262313843,
"learning_rate": 0.00029808403692689953,
"loss": 2.003041076660156,
"step": 4170
},
{
"epoch": 0.53504,
"grad_norm": 0.7704663276672363,
"learning_rate": 0.00029807439086144575,
"loss": 2.008600044250488,
"step": 4180
},
{
"epoch": 0.53632,
"grad_norm": 0.7134334444999695,
"learning_rate": 0.0002980647207318522,
"loss": 2.0230785369873048,
"step": 4190
},
{
"epoch": 0.5376,
"grad_norm": 0.6899425983428955,
"learning_rate": 0.00029805502653969035,
"loss": 2.014421272277832,
"step": 4200
},
{
"epoch": 0.53888,
"grad_norm": 0.7387205362319946,
"learning_rate": 0.0002980453082865357,
"loss": 1.9908321380615235,
"step": 4210
},
{
"epoch": 0.54016,
"grad_norm": 0.6980050802230835,
"learning_rate": 0.0002980355659739676,
"loss": 1.9989233016967773,
"step": 4220
},
{
"epoch": 0.54144,
"grad_norm": 0.6020520925521851,
"learning_rate": 0.00029802579960356933,
"loss": 2.00181884765625,
"step": 4230
},
{
"epoch": 0.54272,
"grad_norm": 0.6786367893218994,
"learning_rate": 0.0002980160091769279,
"loss": 1.9844453811645508,
"step": 4240
},
{
"epoch": 0.544,
"grad_norm": 0.6527010202407837,
"learning_rate": 0.00029800619469563463,
"loss": 1.9895397186279298,
"step": 4250
},
{
"epoch": 0.54528,
"grad_norm": 0.6559035181999207,
"learning_rate": 0.00029799635616128435,
"loss": 2.0083845138549803,
"step": 4260
},
{
"epoch": 0.54656,
"grad_norm": 0.7348132729530334,
"learning_rate": 0.000297986493575476,
"loss": 2.0096080780029295,
"step": 4270
},
{
"epoch": 0.54784,
"grad_norm": 0.6425591111183167,
"learning_rate": 0.00029797660693981235,
"loss": 1.9981698989868164,
"step": 4280
},
{
"epoch": 0.54912,
"grad_norm": 0.6125735640525818,
"learning_rate": 0.0002979666962559002,
"loss": 2.0015987396240233,
"step": 4290
},
{
"epoch": 0.5504,
"grad_norm": 0.6617387533187866,
"learning_rate": 0.0002979567615253501,
"loss": 2.0117170333862306,
"step": 4300
},
{
"epoch": 0.55168,
"grad_norm": 0.8098113536834717,
"learning_rate": 0.00029794680274977664,
"loss": 1.976462173461914,
"step": 4310
},
{
"epoch": 0.55296,
"grad_norm": 0.7238470911979675,
"learning_rate": 0.00029793681993079817,
"loss": 1.9960739135742187,
"step": 4320
},
{
"epoch": 0.55424,
"grad_norm": 0.8663386702537537,
"learning_rate": 0.0002979268130700371,
"loss": 1.9836992263793944,
"step": 4330
},
{
"epoch": 0.55552,
"grad_norm": 0.7530176043510437,
"learning_rate": 0.00029791678216911974,
"loss": 1.993238639831543,
"step": 4340
},
{
"epoch": 0.5568,
"grad_norm": 0.8118612766265869,
"learning_rate": 0.0002979067272296761,
"loss": 2.0035892486572267,
"step": 4350
},
{
"epoch": 0.55808,
"grad_norm": 0.7913742065429688,
"learning_rate": 0.0002978966482533404,
"loss": 2.002780532836914,
"step": 4360
},
{
"epoch": 0.55936,
"grad_norm": 0.694092333316803,
"learning_rate": 0.0002978865452417505,
"loss": 2.0032726287841798,
"step": 4370
},
{
"epoch": 0.56064,
"grad_norm": 0.6410046815872192,
"learning_rate": 0.0002978764181965483,
"loss": 2.001888465881348,
"step": 4380
},
{
"epoch": 0.56192,
"grad_norm": 0.7601751089096069,
"learning_rate": 0.00029786626711937966,
"loss": 2.0106828689575194,
"step": 4390
},
{
"epoch": 0.5632,
"grad_norm": 0.7212942838668823,
"learning_rate": 0.00029785609201189424,
"loss": 2.023110198974609,
"step": 4400
},
{
"epoch": 0.56448,
"grad_norm": 0.7451926469802856,
"learning_rate": 0.0002978458928757456,
"loss": 2.011552619934082,
"step": 4410
},
{
"epoch": 0.56576,
"grad_norm": 0.6781797409057617,
"learning_rate": 0.00029783566971259125,
"loss": 2.0140331268310545,
"step": 4420
},
{
"epoch": 0.56704,
"grad_norm": 0.7447178959846497,
"learning_rate": 0.0002978254225240926,
"loss": 2.0165130615234377,
"step": 4430
},
{
"epoch": 0.56832,
"grad_norm": 0.7245031595230103,
"learning_rate": 0.0002978151513119149,
"loss": 1.9940263748168945,
"step": 4440
},
{
"epoch": 0.5696,
"grad_norm": 0.756145715713501,
"learning_rate": 0.0002978048560777275,
"loss": 1.975267219543457,
"step": 4450
},
{
"epoch": 0.57088,
"grad_norm": 0.7302019000053406,
"learning_rate": 0.00029779453682320345,
"loss": 1.992108154296875,
"step": 4460
},
{
"epoch": 0.57216,
"grad_norm": 0.8035944104194641,
"learning_rate": 0.0002977841935500197,
"loss": 1.9908878326416015,
"step": 4470
},
{
"epoch": 0.57344,
"grad_norm": 0.7327392101287842,
"learning_rate": 0.00029777382625985734,
"loss": 2.000600051879883,
"step": 4480
},
{
"epoch": 0.57472,
"grad_norm": 0.6741060018539429,
"learning_rate": 0.00029776343495440104,
"loss": 2.0095552444458007,
"step": 4490
},
{
"epoch": 0.576,
"grad_norm": 0.6768994331359863,
"learning_rate": 0.00029775301963533957,
"loss": 1.983211135864258,
"step": 4500
},
{
"epoch": 0.57728,
"grad_norm": 0.7285369634628296,
"learning_rate": 0.0002977425803043657,
"loss": 2.011127471923828,
"step": 4510
},
{
"epoch": 0.57856,
"grad_norm": 0.6478345394134521,
"learning_rate": 0.00029773211696317573,
"loss": 1.9803707122802734,
"step": 4520
},
{
"epoch": 0.57984,
"grad_norm": 0.7388439774513245,
"learning_rate": 0.0002977216296134703,
"loss": 1.9949949264526368,
"step": 4530
},
{
"epoch": 0.58112,
"grad_norm": 0.6757745742797852,
"learning_rate": 0.0002977111182569536,
"loss": 2.025724983215332,
"step": 4540
},
{
"epoch": 0.5824,
"grad_norm": 0.6242362856864929,
"learning_rate": 0.00029770058289533397,
"loss": 2.0046735763549806,
"step": 4550
},
{
"epoch": 0.58368,
"grad_norm": 0.6959071159362793,
"learning_rate": 0.0002976900235303235,
"loss": 1.9909477233886719,
"step": 4560
},
{
"epoch": 0.58496,
"grad_norm": 1.1365975141525269,
"learning_rate": 0.0002976794401636383,
"loss": 2.0057315826416016,
"step": 4570
},
{
"epoch": 0.58624,
"grad_norm": 0.6782042384147644,
"learning_rate": 0.00029766883279699826,
"loss": 2.0111324310302736,
"step": 4580
},
{
"epoch": 0.58752,
"grad_norm": 1.024519681930542,
"learning_rate": 0.0002976582014321272,
"loss": 1.995925521850586,
"step": 4590
},
{
"epoch": 0.5888,
"grad_norm": 0.7553693056106567,
"learning_rate": 0.0002976475460707529,
"loss": 2.0124210357666015,
"step": 4600
},
{
"epoch": 0.59008,
"grad_norm": 0.6609204411506653,
"learning_rate": 0.00029763686671460703,
"loss": 1.9881935119628906,
"step": 4610
},
{
"epoch": 0.59136,
"grad_norm": 0.7214299440383911,
"learning_rate": 0.00029762616336542514,
"loss": 1.9801551818847656,
"step": 4620
},
{
"epoch": 0.59264,
"grad_norm": 0.6682612895965576,
"learning_rate": 0.00029761543602494655,
"loss": 2.023514175415039,
"step": 4630
},
{
"epoch": 0.59392,
"grad_norm": 0.7894415855407715,
"learning_rate": 0.0002976046846949148,
"loss": 2.010751724243164,
"step": 4640
},
{
"epoch": 0.5952,
"grad_norm": 0.6574305891990662,
"learning_rate": 0.0002975939093770769,
"loss": 1.976632308959961,
"step": 4650
},
{
"epoch": 0.59648,
"grad_norm": 0.6413716673851013,
"learning_rate": 0.00029758311007318416,
"loss": 2.0151065826416015,
"step": 4660
},
{
"epoch": 0.59776,
"grad_norm": 0.7089874148368835,
"learning_rate": 0.00029757228678499155,
"loss": 1.9980993270874023,
"step": 4670
},
{
"epoch": 0.59904,
"grad_norm": 0.6705209612846375,
"learning_rate": 0.000297561439514258,
"loss": 2.0038993835449217,
"step": 4680
},
{
"epoch": 0.60032,
"grad_norm": 0.6938350796699524,
"learning_rate": 0.00029755056826274637,
"loss": 1.9767511367797852,
"step": 4690
},
{
"epoch": 0.6016,
"grad_norm": 0.733703076839447,
"learning_rate": 0.0002975396730322234,
"loss": 2.0173906326293944,
"step": 4700
},
{
"epoch": 0.60288,
"grad_norm": 0.6944065093994141,
"learning_rate": 0.0002975287538244597,
"loss": 2.012947654724121,
"step": 4710
},
{
"epoch": 0.60416,
"grad_norm": 0.7177563905715942,
"learning_rate": 0.0002975178106412298,
"loss": 1.9930303573608399,
"step": 4720
},
{
"epoch": 0.60544,
"grad_norm": 0.7881021499633789,
"learning_rate": 0.000297506843484312,
"loss": 2.0142608642578126,
"step": 4730
},
{
"epoch": 0.60672,
"grad_norm": 0.7584002614021301,
"learning_rate": 0.00029749585235548887,
"loss": 2.0150390625,
"step": 4740
},
{
"epoch": 0.608,
"grad_norm": 0.7548288106918335,
"learning_rate": 0.00029748483725654636,
"loss": 2.001949691772461,
"step": 4750
},
{
"epoch": 0.60928,
"grad_norm": 0.7352362871170044,
"learning_rate": 0.00029747379818927474,
"loss": 2.014604187011719,
"step": 4760
},
{
"epoch": 0.61056,
"grad_norm": 0.736960768699646,
"learning_rate": 0.00029746273515546793,
"loss": 2.0182096481323244,
"step": 4770
},
{
"epoch": 0.61184,
"grad_norm": 0.8847310543060303,
"learning_rate": 0.00029745164815692387,
"loss": 1.9898075103759765,
"step": 4780
},
{
"epoch": 0.61312,
"grad_norm": 0.7176423072814941,
"learning_rate": 0.0002974405371954443,
"loss": 2.003055953979492,
"step": 4790
},
{
"epoch": 0.6144,
"grad_norm": 0.7516958713531494,
"learning_rate": 0.00029742940227283504,
"loss": 1.9840295791625977,
"step": 4800
},
{
"epoch": 0.61568,
"grad_norm": 0.7542930245399475,
"learning_rate": 0.0002974182433909055,
"loss": 2.0189092636108397,
"step": 4810
},
{
"epoch": 0.61696,
"grad_norm": 0.7811306118965149,
"learning_rate": 0.0002974070605514692,
"loss": 2.005120849609375,
"step": 4820
},
{
"epoch": 0.61824,
"grad_norm": 0.6935572028160095,
"learning_rate": 0.00029739585375634356,
"loss": 2.0136289596557617,
"step": 4830
},
{
"epoch": 0.61952,
"grad_norm": 0.9725347757339478,
"learning_rate": 0.0002973846230073498,
"loss": 2.0010726928710936,
"step": 4840
},
{
"epoch": 0.6208,
"grad_norm": 0.6915906071662903,
"learning_rate": 0.0002973733683063131,
"loss": 2.010584259033203,
"step": 4850
},
{
"epoch": 0.62208,
"grad_norm": 0.771823525428772,
"learning_rate": 0.00029736208965506247,
"loss": 2.0161518096923827,
"step": 4860
},
{
"epoch": 0.62336,
"grad_norm": 0.6337702870368958,
"learning_rate": 0.00029735078705543086,
"loss": 2.0008981704711912,
"step": 4870
},
{
"epoch": 0.62464,
"grad_norm": 0.6666226387023926,
"learning_rate": 0.0002973394605092551,
"loss": 2.010310173034668,
"step": 4880
},
{
"epoch": 0.62592,
"grad_norm": 0.7635229825973511,
"learning_rate": 0.00029732811001837594,
"loss": 2.0058834075927736,
"step": 4890
},
{
"epoch": 0.6272,
"grad_norm": 0.9127666354179382,
"learning_rate": 0.0002973167355846379,
"loss": 2.0224124908447267,
"step": 4900
},
{
"epoch": 0.62848,
"grad_norm": 0.8189308047294617,
"learning_rate": 0.0002973053372098896,
"loss": 2.007943344116211,
"step": 4910
},
{
"epoch": 0.62976,
"grad_norm": 0.7304124236106873,
"learning_rate": 0.0002972939148959834,
"loss": 2.000100517272949,
"step": 4920
},
{
"epoch": 0.63104,
"grad_norm": 0.6603564620018005,
"learning_rate": 0.00029728246864477555,
"loss": 2.0129776000976562,
"step": 4930
},
{
"epoch": 0.63232,
"grad_norm": 0.6280004382133484,
"learning_rate": 0.00029727099845812616,
"loss": 2.0045780181884765,
"step": 4940
},
{
"epoch": 0.6336,
"grad_norm": 0.7431449294090271,
"learning_rate": 0.00029725950433789954,
"loss": 1.98687744140625,
"step": 4950
},
{
"epoch": 0.63488,
"grad_norm": 0.8049905896186829,
"learning_rate": 0.0002972479862859634,
"loss": 2.004106330871582,
"step": 4960
},
{
"epoch": 0.63616,
"grad_norm": 0.7185505628585815,
"learning_rate": 0.00029723644430418956,
"loss": 1.998849868774414,
"step": 4970
},
{
"epoch": 0.63744,
"grad_norm": 0.8346318602561951,
"learning_rate": 0.0002972248783944539,
"loss": 1.9910211563110352,
"step": 4980
},
{
"epoch": 0.63872,
"grad_norm": 0.686622142791748,
"learning_rate": 0.0002972132885586361,
"loss": 1.9930900573730468,
"step": 4990
},
{
"epoch": 0.64,
"grad_norm": 0.7573023438453674,
"learning_rate": 0.00029720167479861946,
"loss": 2.01367244720459,
"step": 5000
},
{
"epoch": 0.64128,
"grad_norm": 0.7065608501434326,
"learning_rate": 0.0002971900371162915,
"loss": 1.9992128372192384,
"step": 5010
},
{
"epoch": 0.64256,
"grad_norm": 0.7289401292800903,
"learning_rate": 0.00029717837551354354,
"loss": 1.9842073440551757,
"step": 5020
},
{
"epoch": 0.64384,
"grad_norm": 0.6458973288536072,
"learning_rate": 0.00029716668999227064,
"loss": 2.0016260147094727,
"step": 5030
},
{
"epoch": 0.64512,
"grad_norm": 0.7114098072052002,
"learning_rate": 0.0002971549805543719,
"loss": 1.9759035110473633,
"step": 5040
},
{
"epoch": 0.6464,
"grad_norm": 0.7502028346061707,
"learning_rate": 0.0002971432472017503,
"loss": 1.9977811813354491,
"step": 5050
},
{
"epoch": 0.64768,
"grad_norm": 0.7050338387489319,
"learning_rate": 0.00029713148993631267,
"loss": 2.0023754119873045,
"step": 5060
},
{
"epoch": 0.64896,
"grad_norm": 0.7351704835891724,
"learning_rate": 0.0002971197087599697,
"loss": 1.9990671157836915,
"step": 5070
},
{
"epoch": 0.65024,
"grad_norm": 0.684573769569397,
"learning_rate": 0.000297107903674636,
"loss": 2.0173965454101563,
"step": 5080
},
{
"epoch": 0.65152,
"grad_norm": 0.7622045278549194,
"learning_rate": 0.0002970960746822301,
"loss": 1.9990699768066407,
"step": 5090
},
{
"epoch": 0.6528,
"grad_norm": 0.5931624174118042,
"learning_rate": 0.00029708422178467426,
"loss": 2.0040538787841795,
"step": 5100
},
{
"epoch": 0.65408,
"grad_norm": 0.6407083868980408,
"learning_rate": 0.0002970723449838949,
"loss": 2.00338191986084,
"step": 5110
},
{
"epoch": 0.65536,
"grad_norm": 0.7045226097106934,
"learning_rate": 0.0002970604442818221,
"loss": 1.996185874938965,
"step": 5120
},
{
"epoch": 0.65664,
"grad_norm": 0.6836894154548645,
"learning_rate": 0.0002970485196803898,
"loss": 1.994027328491211,
"step": 5130
},
{
"epoch": 0.65792,
"grad_norm": 0.6975918412208557,
"learning_rate": 0.000297036571181536,
"loss": 1.9909107208251953,
"step": 5140
},
{
"epoch": 0.6592,
"grad_norm": 0.7193573713302612,
"learning_rate": 0.00029702459878720255,
"loss": 1.9966390609741211,
"step": 5150
},
{
"epoch": 0.66048,
"grad_norm": 0.6120880246162415,
"learning_rate": 0.000297012602499335,
"loss": 1.9817461013793944,
"step": 5160
},
{
"epoch": 0.66176,
"grad_norm": 0.759924054145813,
"learning_rate": 0.000297000582319883,
"loss": 1.9850471496582032,
"step": 5170
},
{
"epoch": 0.66304,
"grad_norm": 0.6980900764465332,
"learning_rate": 0.0002969885382508,
"loss": 1.9931489944458007,
"step": 5180
},
{
"epoch": 0.66432,
"grad_norm": 0.7123715877532959,
"learning_rate": 0.00029697647029404325,
"loss": 2.000169563293457,
"step": 5190
},
{
"epoch": 0.6656,
"grad_norm": 0.6090735197067261,
"learning_rate": 0.000296964378451574,
"loss": 1.9939422607421875,
"step": 5200
},
{
"epoch": 0.66688,
"grad_norm": 0.7080000042915344,
"learning_rate": 0.0002969522627253574,
"loss": 1.9911760330200194,
"step": 5210
},
{
"epoch": 0.66816,
"grad_norm": 0.6564937233924866,
"learning_rate": 0.0002969401231173624,
"loss": 1.9953861236572266,
"step": 5220
},
{
"epoch": 0.66944,
"grad_norm": 0.6857775449752808,
"learning_rate": 0.0002969279596295618,
"loss": 2.0081436157226564,
"step": 5230
},
{
"epoch": 0.67072,
"grad_norm": 0.7370122671127319,
"learning_rate": 0.00029691577226393235,
"loss": 2.0082754135131835,
"step": 5240
},
{
"epoch": 0.672,
"grad_norm": 0.6324231624603271,
"learning_rate": 0.0002969035610224547,
"loss": 2.0048879623413085,
"step": 5250
},
{
"epoch": 0.67328,
"grad_norm": 0.7700392007827759,
"learning_rate": 0.0002968913259071133,
"loss": 2.0072811126708983,
"step": 5260
},
{
"epoch": 0.67456,
"grad_norm": 0.6762933731079102,
"learning_rate": 0.00029687906691989655,
"loss": 1.9954465866088866,
"step": 5270
},
{
"epoch": 0.67584,
"grad_norm": 0.7391619086265564,
"learning_rate": 0.00029686678406279674,
"loss": 1.9988208770751954,
"step": 5280
},
{
"epoch": 0.67712,
"grad_norm": 0.6812599301338196,
"learning_rate": 0.00029685447733780995,
"loss": 1.9795747756958009,
"step": 5290
},
{
"epoch": 0.6784,
"grad_norm": 0.6824685335159302,
"learning_rate": 0.00029684214674693623,
"loss": 2.0002880096435547,
"step": 5300
},
{
"epoch": 0.67968,
"grad_norm": 0.7883182168006897,
"learning_rate": 0.00029682979229217943,
"loss": 1.9892650604248048,
"step": 5310
},
{
"epoch": 0.68096,
"grad_norm": 0.8476542234420776,
"learning_rate": 0.0002968174139755474,
"loss": 2.0148120880126954,
"step": 5320
},
{
"epoch": 0.68224,
"grad_norm": 0.7227822542190552,
"learning_rate": 0.00029680501179905173,
"loss": 2.0036792755126953,
"step": 5330
},
{
"epoch": 0.68352,
"grad_norm": 0.6343380212783813,
"learning_rate": 0.00029679258576470794,
"loss": 2.0134708404541017,
"step": 5340
},
{
"epoch": 0.6848,
"grad_norm": 0.6354977488517761,
"learning_rate": 0.00029678013587453543,
"loss": 2.0034976959228517,
"step": 5350
},
{
"epoch": 0.68608,
"grad_norm": 0.7666685581207275,
"learning_rate": 0.0002967676621305575,
"loss": 1.9847345352172852,
"step": 5360
},
{
"epoch": 0.68736,
"grad_norm": 0.8441287875175476,
"learning_rate": 0.0002967551645348013,
"loss": 2.006691551208496,
"step": 5370
},
{
"epoch": 0.68864,
"grad_norm": 0.6251808404922485,
"learning_rate": 0.0002967426430892979,
"loss": 1.9834270477294922,
"step": 5380
},
{
"epoch": 0.68992,
"grad_norm": 0.636553168296814,
"learning_rate": 0.0002967300977960821,
"loss": 1.9941967010498047,
"step": 5390
},
{
"epoch": 0.6912,
"grad_norm": 0.6793889999389648,
"learning_rate": 0.0002967175286571928,
"loss": 2.016760063171387,
"step": 5400
},
{
"epoch": 0.69248,
"grad_norm": 0.7251984477043152,
"learning_rate": 0.0002967049356746726,
"loss": 1.9951417922973633,
"step": 5410
},
{
"epoch": 0.69376,
"grad_norm": 0.8330734372138977,
"learning_rate": 0.00029669231885056806,
"loss": 1.992695426940918,
"step": 5420
},
{
"epoch": 0.69504,
"grad_norm": 0.6929616928100586,
"learning_rate": 0.00029667967818692957,
"loss": 1.9796766281127929,
"step": 5430
},
{
"epoch": 0.69632,
"grad_norm": 0.73932945728302,
"learning_rate": 0.0002966670136858114,
"loss": 1.9927392959594727,
"step": 5440
},
{
"epoch": 0.6976,
"grad_norm": 0.728484034538269,
"learning_rate": 0.00029665432534927176,
"loss": 1.990020179748535,
"step": 5450
},
{
"epoch": 0.69888,
"grad_norm": 0.8292369842529297,
"learning_rate": 0.0002966416131793726,
"loss": 2.007571220397949,
"step": 5460
},
{
"epoch": 0.70016,
"grad_norm": 0.6958839297294617,
"learning_rate": 0.00029662887717818,
"loss": 1.9895818710327149,
"step": 5470
},
{
"epoch": 0.70144,
"grad_norm": 0.8625398874282837,
"learning_rate": 0.0002966161173477635,
"loss": 2.008391761779785,
"step": 5480
},
{
"epoch": 0.70272,
"grad_norm": 0.7190197110176086,
"learning_rate": 0.00029660333369019683,
"loss": 1.9892658233642577,
"step": 5490
},
{
"epoch": 0.704,
"grad_norm": 0.6254742741584778,
"learning_rate": 0.00029659052620755757,
"loss": 1.9928400039672851,
"step": 5500
},
{
"epoch": 0.70528,
"grad_norm": 0.7220839262008667,
"learning_rate": 0.0002965776949019271,
"loss": 1.9931129455566405,
"step": 5510
},
{
"epoch": 0.70656,
"grad_norm": 0.6875084042549133,
"learning_rate": 0.0002965648397753907,
"loss": 2.006221961975098,
"step": 5520
},
{
"epoch": 0.70784,
"grad_norm": 0.6967015862464905,
"learning_rate": 0.0002965519608300374,
"loss": 2.0144590377807616,
"step": 5530
},
{
"epoch": 0.70912,
"grad_norm": 0.7796663045883179,
"learning_rate": 0.0002965390580679603,
"loss": 1.9761390686035156,
"step": 5540
},
{
"epoch": 0.7104,
"grad_norm": 0.6932560801506042,
"learning_rate": 0.0002965261314912563,
"loss": 1.9977296829223632,
"step": 5550
},
{
"epoch": 0.71168,
"grad_norm": 0.6555649638175964,
"learning_rate": 0.00029651318110202605,
"loss": 1.9888612747192382,
"step": 5560
},
{
"epoch": 0.71296,
"grad_norm": 0.7425674796104431,
"learning_rate": 0.00029650020690237425,
"loss": 1.9898294448852538,
"step": 5570
},
{
"epoch": 0.71424,
"grad_norm": 0.754615068435669,
"learning_rate": 0.0002964872088944093,
"loss": 1.9967521667480468,
"step": 5580
},
{
"epoch": 0.71552,
"grad_norm": 0.6292251944541931,
"learning_rate": 0.0002964741870802437,
"loss": 2.0030208587646485,
"step": 5590
},
{
"epoch": 0.7168,
"grad_norm": 0.7437636256217957,
"learning_rate": 0.0002964611414619936,
"loss": 2.0063127517700194,
"step": 5600
},
{
"epoch": 0.71808,
"grad_norm": 0.6832345128059387,
"learning_rate": 0.000296448072041779,
"loss": 1.9996273040771484,
"step": 5610
},
{
"epoch": 0.71936,
"grad_norm": 0.6542320847511292,
"learning_rate": 0.000296434978821724,
"loss": 1.9821640014648438,
"step": 5620
},
{
"epoch": 0.72064,
"grad_norm": 0.6850648522377014,
"learning_rate": 0.0002964218618039564,
"loss": 1.9940942764282226,
"step": 5630
},
{
"epoch": 0.72192,
"grad_norm": 0.7427983283996582,
"learning_rate": 0.00029640872099060783,
"loss": 1.981001091003418,
"step": 5640
},
{
"epoch": 0.7232,
"grad_norm": 0.7683842182159424,
"learning_rate": 0.0002963955563838139,
"loss": 2.008561897277832,
"step": 5650
},
{
"epoch": 0.72448,
"grad_norm": 0.6730973124504089,
"learning_rate": 0.00029638236798571404,
"loss": 1.996862030029297,
"step": 5660
},
{
"epoch": 0.72576,
"grad_norm": 0.6657568216323853,
"learning_rate": 0.00029636915579845155,
"loss": 1.9981700897216796,
"step": 5670
},
{
"epoch": 0.72704,
"grad_norm": 0.7134661674499512,
"learning_rate": 0.0002963559198241736,
"loss": 1.9864439010620116,
"step": 5680
},
{
"epoch": 0.72832,
"grad_norm": 0.7284188866615295,
"learning_rate": 0.00029634266006503123,
"loss": 2.0042444229125977,
"step": 5690
},
{
"epoch": 0.7296,
"grad_norm": 0.7229200005531311,
"learning_rate": 0.0002963293765231793,
"loss": 1.9872091293334961,
"step": 5700
},
{
"epoch": 0.73088,
"grad_norm": 0.6603932976722717,
"learning_rate": 0.00029631606920077654,
"loss": 2.0120763778686523,
"step": 5710
},
{
"epoch": 0.73216,
"grad_norm": 0.8360553979873657,
"learning_rate": 0.00029630273809998567,
"loss": 1.9983232498168946,
"step": 5720
},
{
"epoch": 0.73344,
"grad_norm": 0.611983060836792,
"learning_rate": 0.0002962893832229731,
"loss": 1.9799180984497071,
"step": 5730
},
{
"epoch": 0.73472,
"grad_norm": 0.6826144456863403,
"learning_rate": 0.0002962760045719092,
"loss": 1.985218620300293,
"step": 5740
},
{
"epoch": 0.736,
"grad_norm": 0.6724496483802795,
"learning_rate": 0.0002962626021489682,
"loss": 1.9969600677490233,
"step": 5750
},
{
"epoch": 0.73728,
"grad_norm": 0.6387225389480591,
"learning_rate": 0.0002962491759563282,
"loss": 2.0010164260864256,
"step": 5760
},
{
"epoch": 0.73856,
"grad_norm": 0.7522284388542175,
"learning_rate": 0.00029623572599617106,
"loss": 2.0010658264160157,
"step": 5770
},
{
"epoch": 0.73984,
"grad_norm": 0.6354997158050537,
"learning_rate": 0.0002962222522706827,
"loss": 2.015804481506348,
"step": 5780
},
{
"epoch": 0.74112,
"grad_norm": 0.671524703502655,
"learning_rate": 0.00029620875478205273,
"loss": 2.0011077880859376,
"step": 5790
},
{
"epoch": 0.7424,
"grad_norm": 0.6422104835510254,
"learning_rate": 0.00029619523353247463,
"loss": 1.9921817779541016,
"step": 5800
},
{
"epoch": 0.74368,
"grad_norm": 0.6711571216583252,
"learning_rate": 0.0002961816885241458,
"loss": 1.9803316116333007,
"step": 5810
},
{
"epoch": 0.74496,
"grad_norm": 0.8022178411483765,
"learning_rate": 0.00029616811975926763,
"loss": 1.9788673400878907,
"step": 5820
},
{
"epoch": 0.74624,
"grad_norm": 0.6730902194976807,
"learning_rate": 0.0002961545272400451,
"loss": 1.9903987884521483,
"step": 5830
},
{
"epoch": 0.74752,
"grad_norm": 0.7900033593177795,
"learning_rate": 0.0002961409109686872,
"loss": 2.013372802734375,
"step": 5840
},
{
"epoch": 0.7488,
"grad_norm": 0.6188972592353821,
"learning_rate": 0.00029612727094740674,
"loss": 1.9988031387329102,
"step": 5850
},
{
"epoch": 0.75008,
"grad_norm": 0.6830716729164124,
"learning_rate": 0.00029611360717842047,
"loss": 1.9980751037597657,
"step": 5860
},
{
"epoch": 0.75136,
"grad_norm": 0.7211880683898926,
"learning_rate": 0.00029609991966394897,
"loss": 2.0045663833618166,
"step": 5870
},
{
"epoch": 0.75264,
"grad_norm": 0.6811301708221436,
"learning_rate": 0.0002960862084062166,
"loss": 2.0147485733032227,
"step": 5880
},
{
"epoch": 0.75392,
"grad_norm": 0.6604957580566406,
"learning_rate": 0.0002960724734074516,
"loss": 1.987942886352539,
"step": 5890
},
{
"epoch": 0.7552,
"grad_norm": 0.7545401453971863,
"learning_rate": 0.0002960587146698861,
"loss": 2.0053482055664062,
"step": 5900
},
{
"epoch": 0.75648,
"grad_norm": 0.6331950426101685,
"learning_rate": 0.00029604493219575616,
"loss": 1.9871101379394531,
"step": 5910
},
{
"epoch": 0.75776,
"grad_norm": 0.6771983504295349,
"learning_rate": 0.0002960311259873015,
"loss": 1.9838722229003907,
"step": 5920
},
{
"epoch": 0.75904,
"grad_norm": 0.7738776206970215,
"learning_rate": 0.000296017296046766,
"loss": 2.0045019149780274,
"step": 5930
},
{
"epoch": 0.76032,
"grad_norm": 0.7317860722541809,
"learning_rate": 0.0002960034423763971,
"loss": 2.015328788757324,
"step": 5940
},
{
"epoch": 0.7616,
"grad_norm": 0.6838796138763428,
"learning_rate": 0.00029598956497844616,
"loss": 1.993131446838379,
"step": 5950
},
{
"epoch": 0.76288,
"grad_norm": 0.7081223130226135,
"learning_rate": 0.00029597566385516856,
"loss": 1.9932073593139648,
"step": 5960
},
{
"epoch": 0.76416,
"grad_norm": 0.7130403518676758,
"learning_rate": 0.0002959617390088234,
"loss": 1.995915985107422,
"step": 5970
},
{
"epoch": 0.76544,
"grad_norm": 0.6577734351158142,
"learning_rate": 0.0002959477904416736,
"loss": 2.001402473449707,
"step": 5980
},
{
"epoch": 0.76672,
"grad_norm": 0.8433389663696289,
"learning_rate": 0.00029593381815598606,
"loss": 1.9908720016479493,
"step": 5990
},
{
"epoch": 0.768,
"grad_norm": 0.7578377723693848,
"learning_rate": 0.00029591982215403144,
"loss": 1.9951087951660156,
"step": 6000
},
{
"epoch": 0.76928,
"grad_norm": 0.6299430131912231,
"learning_rate": 0.0002959058024380843,
"loss": 1.9930356979370116,
"step": 6010
},
{
"epoch": 0.77056,
"grad_norm": 0.7663775682449341,
"learning_rate": 0.00029589175901042304,
"loss": 1.994969940185547,
"step": 6020
},
{
"epoch": 0.77184,
"grad_norm": 0.7576585412025452,
"learning_rate": 0.0002958776918733299,
"loss": 2.008650779724121,
"step": 6030
},
{
"epoch": 0.77312,
"grad_norm": 1.5998573303222656,
"learning_rate": 0.00029586360102909097,
"loss": 1.9937719345092773,
"step": 6040
},
{
"epoch": 0.7744,
"grad_norm": 0.8406224250793457,
"learning_rate": 0.0002958494864799962,
"loss": 2.0037378311157226,
"step": 6050
},
{
"epoch": 0.77568,
"grad_norm": 0.7180790901184082,
"learning_rate": 0.00029583534822833944,
"loss": 2.002363586425781,
"step": 6060
},
{
"epoch": 0.77696,
"grad_norm": 0.604666531085968,
"learning_rate": 0.0002958211862764183,
"loss": 1.9915838241577148,
"step": 6070
},
{
"epoch": 0.77824,
"grad_norm": 0.6440698504447937,
"learning_rate": 0.0002958070006265344,
"loss": 1.9986055374145508,
"step": 6080
},
{
"epoch": 0.77952,
"grad_norm": 0.6588718891143799,
"learning_rate": 0.000295792791280993,
"loss": 1.9881462097167968,
"step": 6090
},
{
"epoch": 0.7808,
"grad_norm": 0.7286064624786377,
"learning_rate": 0.0002957785582421034,
"loss": 2.000356674194336,
"step": 6100
},
{
"epoch": 0.78208,
"grad_norm": 0.8021141886711121,
"learning_rate": 0.0002957643015121785,
"loss": 1.9829841613769532,
"step": 6110
},
{
"epoch": 0.78336,
"grad_norm": 0.7439085245132446,
"learning_rate": 0.0002957500210935354,
"loss": 1.988357162475586,
"step": 6120
},
{
"epoch": 0.78464,
"grad_norm": 0.6714260578155518,
"learning_rate": 0.00029573571698849477,
"loss": 2.0017559051513674,
"step": 6130
},
{
"epoch": 0.78592,
"grad_norm": 0.654412031173706,
"learning_rate": 0.0002957213891993813,
"loss": 1.9880069732666015,
"step": 6140
},
{
"epoch": 0.7872,
"grad_norm": 0.6551602482795715,
"learning_rate": 0.00029570703772852336,
"loss": 2.0161787033081056,
"step": 6150
},
{
"epoch": 0.78848,
"grad_norm": 0.6666812300682068,
"learning_rate": 0.00029569266257825335,
"loss": 1.9900634765625,
"step": 6160
},
{
"epoch": 0.78976,
"grad_norm": 0.7876213192939758,
"learning_rate": 0.00029567826375090736,
"loss": 2.003603935241699,
"step": 6170
},
{
"epoch": 0.79104,
"grad_norm": 1.2557580471038818,
"learning_rate": 0.0002956638412488254,
"loss": 1.9852615356445313,
"step": 6180
},
{
"epoch": 0.79232,
"grad_norm": 0.6098614931106567,
"learning_rate": 0.0002956493950743514,
"loss": 1.9911855697631835,
"step": 6190
},
{
"epoch": 0.7936,
"grad_norm": 0.7123518586158752,
"learning_rate": 0.00029563492522983304,
"loss": 1.9931888580322266,
"step": 6200
},
{
"epoch": 0.79488,
"grad_norm": 0.6955991983413696,
"learning_rate": 0.0002956204317176218,
"loss": 2.0002986907958986,
"step": 6210
},
{
"epoch": 0.79616,
"grad_norm": 0.6723616123199463,
"learning_rate": 0.0002956059145400731,
"loss": 2.0050445556640626,
"step": 6220
},
{
"epoch": 0.79744,
"grad_norm": 0.7482268810272217,
"learning_rate": 0.0002955913736995463,
"loss": 1.996047592163086,
"step": 6230
},
{
"epoch": 0.79872,
"grad_norm": 0.6395532488822937,
"learning_rate": 0.00029557680919840433,
"loss": 2.012595558166504,
"step": 6240
},
{
"epoch": 0.8,
"grad_norm": 0.6754005551338196,
"learning_rate": 0.00029556222103901416,
"loss": 1.9906475067138671,
"step": 6250
},
{
"epoch": 0.80128,
"grad_norm": 0.6359081864356995,
"learning_rate": 0.00029554760922374665,
"loss": 1.979934310913086,
"step": 6260
},
{
"epoch": 0.80256,
"grad_norm": 0.5925866365432739,
"learning_rate": 0.0002955329737549764,
"loss": 2.0107406616210937,
"step": 6270
},
{
"epoch": 0.80384,
"grad_norm": 0.7143832445144653,
"learning_rate": 0.00029551831463508175,
"loss": 1.9844533920288085,
"step": 6280
},
{
"epoch": 0.80512,
"grad_norm": 0.7044582366943359,
"learning_rate": 0.0002955036318664452,
"loss": 1.9771629333496095,
"step": 6290
},
{
"epoch": 0.8064,
"grad_norm": 0.8326923847198486,
"learning_rate": 0.0002954889254514527,
"loss": 1.9971778869628907,
"step": 6300
},
{
"epoch": 0.80768,
"grad_norm": 0.682141900062561,
"learning_rate": 0.00029547419539249444,
"loss": 1.9877017974853515,
"step": 6310
},
{
"epoch": 0.80896,
"grad_norm": 0.664592981338501,
"learning_rate": 0.0002954594416919642,
"loss": 1.9961456298828124,
"step": 6320
},
{
"epoch": 0.81024,
"grad_norm": 0.6965521574020386,
"learning_rate": 0.0002954446643522596,
"loss": 1.9998743057250976,
"step": 6330
},
{
"epoch": 0.81152,
"grad_norm": 0.6839595437049866,
"learning_rate": 0.00029542986337578216,
"loss": 1.9827798843383788,
"step": 6340
},
{
"epoch": 0.8128,
"grad_norm": 0.7168360948562622,
"learning_rate": 0.0002954150387649373,
"loss": 1.9969907760620118,
"step": 6350
},
{
"epoch": 0.81408,
"grad_norm": 0.7123169302940369,
"learning_rate": 0.0002954001905221342,
"loss": 1.9951913833618165,
"step": 6360
},
{
"epoch": 0.81536,
"grad_norm": 0.6068657040596008,
"learning_rate": 0.0002953853186497859,
"loss": 2.000649642944336,
"step": 6370
},
{
"epoch": 0.81664,
"grad_norm": 0.6289076209068298,
"learning_rate": 0.0002953704231503093,
"loss": 2.0178243637084963,
"step": 6380
},
{
"epoch": 0.81792,
"grad_norm": 0.6308927536010742,
"learning_rate": 0.00029535550402612517,
"loss": 1.9915205001831056,
"step": 6390
},
{
"epoch": 0.8192,
"grad_norm": 0.616354763507843,
"learning_rate": 0.00029534056127965795,
"loss": 2.008121299743652,
"step": 6400
},
{
"epoch": 0.82048,
"grad_norm": 0.8025171756744385,
"learning_rate": 0.0002953255949133362,
"loss": 1.9940040588378907,
"step": 6410
},
{
"epoch": 0.82176,
"grad_norm": 0.5960074663162231,
"learning_rate": 0.00029531060492959206,
"loss": 2.007202911376953,
"step": 6420
},
{
"epoch": 0.82304,
"grad_norm": 0.7320867776870728,
"learning_rate": 0.0002952955913308616,
"loss": 2.001862144470215,
"step": 6430
},
{
"epoch": 0.82432,
"grad_norm": 0.6877573132514954,
"learning_rate": 0.0002952805541195848,
"loss": 1.9913444519042969,
"step": 6440
},
{
"epoch": 0.8256,
"grad_norm": 0.7313185334205627,
"learning_rate": 0.0002952654932982053,
"loss": 2.0005752563476564,
"step": 6450
},
{
"epoch": 0.82688,
"grad_norm": 0.6951473951339722,
"learning_rate": 0.00029525040886917083,
"loss": 1.9982521057128906,
"step": 6460
},
{
"epoch": 0.82816,
"grad_norm": 0.6680341362953186,
"learning_rate": 0.0002952353008349328,
"loss": 2.009914779663086,
"step": 6470
},
{
"epoch": 0.82944,
"grad_norm": 0.7091737389564514,
"learning_rate": 0.0002952201691979464,
"loss": 1.9832798004150392,
"step": 6480
},
{
"epoch": 0.83072,
"grad_norm": 0.7563579678535461,
"learning_rate": 0.00029520501396067077,
"loss": 1.9979364395141601,
"step": 6490
},
{
"epoch": 0.832,
"grad_norm": 0.6675616502761841,
"learning_rate": 0.00029518983512556886,
"loss": 2.0016298294067383,
"step": 6500
},
{
"epoch": 0.83328,
"grad_norm": 0.7811965346336365,
"learning_rate": 0.0002951746326951074,
"loss": 2.0019750595092773,
"step": 6510
},
{
"epoch": 0.83456,
"grad_norm": 0.6802430748939514,
"learning_rate": 0.00029515940667175706,
"loss": 1.9861677169799805,
"step": 6520
},
{
"epoch": 0.83584,
"grad_norm": 2.2662882804870605,
"learning_rate": 0.0002951441570579922,
"loss": 1.9777637481689454,
"step": 6530
},
{
"epoch": 0.83712,
"grad_norm": 0.6255841851234436,
"learning_rate": 0.0002951288838562912,
"loss": 1.9651605606079101,
"step": 6540
},
{
"epoch": 0.8384,
"grad_norm": 0.7399203777313232,
"learning_rate": 0.00029511358706913604,
"loss": 2.0162445068359376,
"step": 6550
},
{
"epoch": 0.83968,
"grad_norm": 0.6955289244651794,
"learning_rate": 0.0002950982666990127,
"loss": 2.0194833755493162,
"step": 6560
},
{
"epoch": 0.84096,
"grad_norm": 0.8125286102294922,
"learning_rate": 0.00029508292274841105,
"loss": 1.9961944580078126,
"step": 6570
},
{
"epoch": 0.84224,
"grad_norm": 0.6711224317550659,
"learning_rate": 0.00029506755521982453,
"loss": 1.9958179473876954,
"step": 6580
},
{
"epoch": 0.84352,
"grad_norm": 0.7141205668449402,
"learning_rate": 0.00029505216411575074,
"loss": 1.9968080520629883,
"step": 6590
},
{
"epoch": 0.8448,
"grad_norm": 0.6302902698516846,
"learning_rate": 0.0002950367494386909,
"loss": 1.9921850204467773,
"step": 6600
},
{
"epoch": 0.84608,
"grad_norm": 0.6486519575119019,
"learning_rate": 0.00029502131119114996,
"loss": 1.9878908157348634,
"step": 6610
},
{
"epoch": 0.84736,
"grad_norm": 0.6470666527748108,
"learning_rate": 0.00029500584937563703,
"loss": 2.011512565612793,
"step": 6620
},
{
"epoch": 0.84864,
"grad_norm": 0.6338657140731812,
"learning_rate": 0.0002949903639946649,
"loss": 2.005690574645996,
"step": 6630
},
{
"epoch": 0.84992,
"grad_norm": 0.7112775444984436,
"learning_rate": 0.00029497485505074997,
"loss": 1.9783432006835937,
"step": 6640
},
{
"epoch": 0.8512,
"grad_norm": 0.7458258271217346,
"learning_rate": 0.0002949593225464128,
"loss": 1.9891368865966796,
"step": 6650
},
{
"epoch": 0.85248,
"grad_norm": 0.7250550389289856,
"learning_rate": 0.0002949437664841775,
"loss": 2.01182861328125,
"step": 6660
},
{
"epoch": 0.85376,
"grad_norm": 0.7721503973007202,
"learning_rate": 0.0002949281868665723,
"loss": 1.9848461151123047,
"step": 6670
},
{
"epoch": 0.85504,
"grad_norm": 0.7179070115089417,
"learning_rate": 0.00029491258369612907,
"loss": 2.006007194519043,
"step": 6680
},
{
"epoch": 0.85632,
"grad_norm": 0.7209992408752441,
"learning_rate": 0.0002948969569753835,
"loss": 2.0215726852416993,
"step": 6690
},
{
"epoch": 0.8576,
"grad_norm": 0.6615075469017029,
"learning_rate": 0.00029488130670687516,
"loss": 1.9979560852050782,
"step": 6700
},
{
"epoch": 0.85888,
"grad_norm": 0.6356847286224365,
"learning_rate": 0.00029486563289314745,
"loss": 1.9941680908203125,
"step": 6710
},
{
"epoch": 0.86016,
"grad_norm": 0.6343762874603271,
"learning_rate": 0.00029484993553674754,
"loss": 1.9781364440917968,
"step": 6720
},
{
"epoch": 0.86144,
"grad_norm": 0.6580159664154053,
"learning_rate": 0.0002948342146402265,
"loss": 2.006149673461914,
"step": 6730
},
{
"epoch": 0.86272,
"grad_norm": 0.6490921974182129,
"learning_rate": 0.00029481847020613924,
"loss": 1.991459274291992,
"step": 6740
},
{
"epoch": 0.864,
"grad_norm": 0.6659132838249207,
"learning_rate": 0.0002948027022370443,
"loss": 1.977961540222168,
"step": 6750
},
{
"epoch": 0.86528,
"grad_norm": 0.6164100170135498,
"learning_rate": 0.0002947869107355044,
"loss": 2.012508773803711,
"step": 6760
},
{
"epoch": 0.86656,
"grad_norm": 0.6852055191993713,
"learning_rate": 0.0002947710957040857,
"loss": 1.9905057907104493,
"step": 6770
},
{
"epoch": 0.86784,
"grad_norm": 0.6740121841430664,
"learning_rate": 0.00029475525714535845,
"loss": 2.011355400085449,
"step": 6780
},
{
"epoch": 0.86912,
"grad_norm": 0.6753290891647339,
"learning_rate": 0.0002947393950618967,
"loss": 1.9824066162109375,
"step": 6790
},
{
"epoch": 0.8704,
"grad_norm": 0.6230936646461487,
"learning_rate": 0.0002947235094562781,
"loss": 1.9923259735107421,
"step": 6800
},
{
"epoch": 0.87168,
"grad_norm": 0.6286994218826294,
"learning_rate": 0.00029470760033108433,
"loss": 1.9875059127807617,
"step": 6810
},
{
"epoch": 0.87296,
"grad_norm": 0.8378221988677979,
"learning_rate": 0.0002946916676889009,
"loss": 1.992207145690918,
"step": 6820
},
{
"epoch": 0.87424,
"grad_norm": 0.6988818645477295,
"learning_rate": 0.00029467571153231705,
"loss": 1.9979055404663086,
"step": 6830
},
{
"epoch": 0.87552,
"grad_norm": 0.6570968627929688,
"learning_rate": 0.0002946597318639259,
"loss": 1.9981571197509767,
"step": 6840
},
{
"epoch": 0.8768,
"grad_norm": 0.6940664052963257,
"learning_rate": 0.0002946437286863243,
"loss": 1.9830677032470703,
"step": 6850
},
{
"epoch": 0.87808,
"grad_norm": 0.7300937175750732,
"learning_rate": 0.0002946277020021131,
"loss": 1.9862096786499024,
"step": 6860
},
{
"epoch": 0.87936,
"grad_norm": 0.7412726283073425,
"learning_rate": 0.0002946116518138968,
"loss": 1.99698486328125,
"step": 6870
},
{
"epoch": 0.88064,
"grad_norm": 0.71746426820755,
"learning_rate": 0.0002945955781242838,
"loss": 2.0053390502929687,
"step": 6880
},
{
"epoch": 0.88192,
"grad_norm": 0.6349291801452637,
"learning_rate": 0.00029457948093588617,
"loss": 1.999542236328125,
"step": 6890
},
{
"epoch": 0.8832,
"grad_norm": 0.654897153377533,
"learning_rate": 0.0002945633602513201,
"loss": 1.995721435546875,
"step": 6900
},
{
"epoch": 0.88448,
"grad_norm": 0.7566924691200256,
"learning_rate": 0.0002945472160732053,
"loss": 1.9950716018676757,
"step": 6910
},
{
"epoch": 0.88576,
"grad_norm": 0.7700677514076233,
"learning_rate": 0.00029453104840416555,
"loss": 1.9884449005126954,
"step": 6920
},
{
"epoch": 0.88704,
"grad_norm": 0.6426023244857788,
"learning_rate": 0.0002945148572468282,
"loss": 1.9813749313354492,
"step": 6930
},
{
"epoch": 0.88832,
"grad_norm": 0.6777772307395935,
"learning_rate": 0.00029449864260382456,
"loss": 1.9967033386230468,
"step": 6940
},
{
"epoch": 0.8896,
"grad_norm": 0.6357970237731934,
"learning_rate": 0.0002944824044777898,
"loss": 1.9828590393066405,
"step": 6950
},
{
"epoch": 0.89088,
"grad_norm": 0.7022501230239868,
"learning_rate": 0.00029446614287136275,
"loss": 1.9888254165649415,
"step": 6960
},
{
"epoch": 0.89216,
"grad_norm": 0.7219241857528687,
"learning_rate": 0.0002944498577871862,
"loss": 2.0122438430786134,
"step": 6970
},
{
"epoch": 0.89344,
"grad_norm": 0.9110201597213745,
"learning_rate": 0.00029443354922790673,
"loss": 2.006223678588867,
"step": 6980
},
{
"epoch": 0.89472,
"grad_norm": 0.6884856224060059,
"learning_rate": 0.00029441721719617467,
"loss": 1.995973777770996,
"step": 6990
},
{
"epoch": 0.896,
"grad_norm": 0.645409882068634,
"learning_rate": 0.0002944008616946442,
"loss": 2.0004106521606446,
"step": 7000
},
{
"epoch": 0.89728,
"grad_norm": 0.711455762386322,
"learning_rate": 0.00029438448272597325,
"loss": 1.9932981491088868,
"step": 7010
},
{
"epoch": 0.89856,
"grad_norm": 0.7177769541740417,
"learning_rate": 0.00029436808029282375,
"loss": 1.997603988647461,
"step": 7020
},
{
"epoch": 0.89984,
"grad_norm": 0.6674689054489136,
"learning_rate": 0.0002943516543978613,
"loss": 1.993707275390625,
"step": 7030
},
{
"epoch": 0.90112,
"grad_norm": 0.6319096684455872,
"learning_rate": 0.00029433520504375526,
"loss": 1.9995529174804687,
"step": 7040
},
{
"epoch": 0.9024,
"grad_norm": 0.709757387638092,
"learning_rate": 0.00029431873223317894,
"loss": 2.0230655670166016,
"step": 7050
},
{
"epoch": 0.90368,
"grad_norm": 0.6909634470939636,
"learning_rate": 0.00029430223596880935,
"loss": 1.9975494384765624,
"step": 7060
},
{
"epoch": 0.90496,
"grad_norm": 0.7052473425865173,
"learning_rate": 0.00029428571625332746,
"loss": 1.9993263244628907,
"step": 7070
},
{
"epoch": 0.90624,
"grad_norm": 0.7152456641197205,
"learning_rate": 0.00029426917308941785,
"loss": 1.9970455169677734,
"step": 7080
},
{
"epoch": 0.90752,
"grad_norm": 0.6790175437927246,
"learning_rate": 0.0002942526064797691,
"loss": 2.0116363525390626,
"step": 7090
},
{
"epoch": 0.9088,
"grad_norm": 0.7606917023658752,
"learning_rate": 0.0002942360164270734,
"loss": 1.9900932312011719,
"step": 7100
},
{
"epoch": 0.91008,
"grad_norm": 0.7293239235877991,
"learning_rate": 0.00029421940293402693,
"loss": 2.0022926330566406,
"step": 7110
},
{
"epoch": 0.91136,
"grad_norm": 0.6462413668632507,
"learning_rate": 0.00029420276600332964,
"loss": 1.998992919921875,
"step": 7120
},
{
"epoch": 0.91264,
"grad_norm": 0.686337947845459,
"learning_rate": 0.0002941861056376853,
"loss": 1.9897741317749023,
"step": 7130
},
{
"epoch": 0.91392,
"grad_norm": 0.6641421914100647,
"learning_rate": 0.00029416942183980126,
"loss": 1.9985929489135743,
"step": 7140
},
{
"epoch": 0.9152,
"grad_norm": 0.7737864851951599,
"learning_rate": 0.000294152714612389,
"loss": 1.9727941513061524,
"step": 7150
},
{
"epoch": 0.91648,
"grad_norm": 0.64877849817276,
"learning_rate": 0.00029413598395816376,
"loss": 2.022666358947754,
"step": 7160
},
{
"epoch": 0.91776,
"grad_norm": 0.6656960248947144,
"learning_rate": 0.0002941192298798443,
"loss": 2.000381851196289,
"step": 7170
},
{
"epoch": 0.91904,
"grad_norm": 0.7156676650047302,
"learning_rate": 0.0002941024523801536,
"loss": 2.003289222717285,
"step": 7180
},
{
"epoch": 0.92032,
"grad_norm": 0.8479704260826111,
"learning_rate": 0.00029408565146181806,
"loss": 1.9910587310791015,
"step": 7190
},
{
"epoch": 0.9216,
"grad_norm": 0.6156359314918518,
"learning_rate": 0.0002940688271275682,
"loss": 1.978067970275879,
"step": 7200
},
{
"epoch": 0.92288,
"grad_norm": 0.6154444813728333,
"learning_rate": 0.0002940519793801381,
"loss": 1.9935375213623048,
"step": 7210
},
{
"epoch": 0.92416,
"grad_norm": 0.6085964441299438,
"learning_rate": 0.0002940351082222658,
"loss": 1.9991844177246094,
"step": 7220
},
{
"epoch": 0.92544,
"grad_norm": 0.5994155406951904,
"learning_rate": 0.0002940182136566931,
"loss": 1.9809038162231445,
"step": 7230
},
{
"epoch": 0.92672,
"grad_norm": 0.6245939135551453,
"learning_rate": 0.0002940012956861656,
"loss": 2.0163652420043947,
"step": 7240
},
{
"epoch": 0.928,
"grad_norm": 0.6001088619232178,
"learning_rate": 0.0002939843543134327,
"loss": 1.9949073791503906,
"step": 7250
},
{
"epoch": 0.92928,
"grad_norm": 0.6890986561775208,
"learning_rate": 0.00029396738954124763,
"loss": 1.9710325241088866,
"step": 7260
},
{
"epoch": 0.93056,
"grad_norm": 0.6825648546218872,
"learning_rate": 0.0002939504013723673,
"loss": 1.9972740173339845,
"step": 7270
},
{
"epoch": 0.93184,
"grad_norm": 0.6806229948997498,
"learning_rate": 0.0002939333898095527,
"loss": 2.008091354370117,
"step": 7280
},
{
"epoch": 0.93312,
"grad_norm": 0.6244903206825256,
"learning_rate": 0.0002939163548555683,
"loss": 2.005999755859375,
"step": 7290
},
{
"epoch": 0.9344,
"grad_norm": 0.6636385917663574,
"learning_rate": 0.0002938992965131825,
"loss": 1.9983814239501954,
"step": 7300
},
{
"epoch": 0.93568,
"grad_norm": 0.6596930623054504,
"learning_rate": 0.00029388221478516765,
"loss": 1.9942008972167968,
"step": 7310
},
{
"epoch": 0.93696,
"grad_norm": 0.6415631175041199,
"learning_rate": 0.00029386510967429973,
"loss": 1.996099090576172,
"step": 7320
},
{
"epoch": 0.93824,
"grad_norm": 0.616016685962677,
"learning_rate": 0.0002938479811833585,
"loss": 1.974690818786621,
"step": 7330
},
{
"epoch": 0.93952,
"grad_norm": 0.7595481872558594,
"learning_rate": 0.00029383082931512755,
"loss": 2.010562515258789,
"step": 7340
},
{
"epoch": 0.9408,
"grad_norm": 0.7548149228096008,
"learning_rate": 0.00029381365407239436,
"loss": 2.013906478881836,
"step": 7350
},
{
"epoch": 0.94208,
"grad_norm": 0.6517251133918762,
"learning_rate": 0.0002937964554579502,
"loss": 1.99677734375,
"step": 7360
},
{
"epoch": 0.94336,
"grad_norm": 0.6647694706916809,
"learning_rate": 0.00029377923347458996,
"loss": 1.9947694778442382,
"step": 7370
},
{
"epoch": 0.94464,
"grad_norm": 0.6167952418327332,
"learning_rate": 0.00029376198812511246,
"loss": 1.9989194869995117,
"step": 7380
},
{
"epoch": 0.94592,
"grad_norm": 0.6266070008277893,
"learning_rate": 0.00029374471941232043,
"loss": 2.007636070251465,
"step": 7390
},
{
"epoch": 0.9472,
"grad_norm": 0.675845742225647,
"learning_rate": 0.00029372742733902016,
"loss": 2.0036916732788086,
"step": 7400
},
{
"epoch": 0.94848,
"grad_norm": 0.674096405506134,
"learning_rate": 0.0002937101119080219,
"loss": 1.9969058990478517,
"step": 7410
},
{
"epoch": 0.94976,
"grad_norm": 0.7183719873428345,
"learning_rate": 0.0002936927731221397,
"loss": 1.9890649795532227,
"step": 7420
},
{
"epoch": 0.95104,
"grad_norm": 0.7150754928588867,
"learning_rate": 0.0002936754109841912,
"loss": 1.9729717254638672,
"step": 7430
},
{
"epoch": 0.95232,
"grad_norm": 0.7129974961280823,
"learning_rate": 0.0002936580254969981,
"loss": 1.9953811645507813,
"step": 7440
},
{
"epoch": 0.9536,
"grad_norm": 0.6096296906471252,
"learning_rate": 0.0002936406166633858,
"loss": 2.001888465881348,
"step": 7450
},
{
"epoch": 0.95488,
"grad_norm": 0.6857330203056335,
"learning_rate": 0.0002936231844861834,
"loss": 1.9927192687988282,
"step": 7460
},
{
"epoch": 0.95616,
"grad_norm": 0.6103941798210144,
"learning_rate": 0.0002936057289682239,
"loss": 1.9911823272705078,
"step": 7470
},
{
"epoch": 0.95744,
"grad_norm": 0.597206175327301,
"learning_rate": 0.0002935882501123441,
"loss": 1.988555335998535,
"step": 7480
},
{
"epoch": 0.95872,
"grad_norm": 0.7180148959159851,
"learning_rate": 0.0002935707479213845,
"loss": 1.978421974182129,
"step": 7490
},
{
"epoch": 0.96,
"grad_norm": 0.6061798930168152,
"learning_rate": 0.0002935532223981895,
"loss": 2.000789833068848,
"step": 7500
},
{
"epoch": 0.96128,
"grad_norm": 0.6952294707298279,
"learning_rate": 0.0002935356735456072,
"loss": 2.0071075439453123,
"step": 7510
},
{
"epoch": 0.96256,
"grad_norm": 0.6120730042457581,
"learning_rate": 0.0002935181013664895,
"loss": 2.0083288192749023,
"step": 7520
},
{
"epoch": 0.96384,
"grad_norm": 0.7803784608840942,
"learning_rate": 0.0002935005058636923,
"loss": 2.009001541137695,
"step": 7530
},
{
"epoch": 0.96512,
"grad_norm": 0.681745171546936,
"learning_rate": 0.0002934828870400749,
"loss": 1.9958309173583983,
"step": 7540
},
{
"epoch": 0.9664,
"grad_norm": 0.7455606460571289,
"learning_rate": 0.0002934652448985007,
"loss": 2.0131885528564455,
"step": 7550
},
{
"epoch": 0.96768,
"grad_norm": 0.7167068719863892,
"learning_rate": 0.0002934475794418368,
"loss": 1.995752716064453,
"step": 7560
},
{
"epoch": 0.96896,
"grad_norm": 0.6225274205207825,
"learning_rate": 0.000293429890672954,
"loss": 1.9854461669921875,
"step": 7570
},
{
"epoch": 0.97024,
"grad_norm": 0.6639164686203003,
"learning_rate": 0.00029341217859472706,
"loss": 1.978256607055664,
"step": 7580
},
{
"epoch": 0.97152,
"grad_norm": 0.6081394553184509,
"learning_rate": 0.0002933944432100344,
"loss": 1.9813619613647462,
"step": 7590
},
{
"epoch": 0.9728,
"grad_norm": 0.6417889595031738,
"learning_rate": 0.00029337668452175826,
"loss": 1.9765779495239257,
"step": 7600
},
{
"epoch": 0.97408,
"grad_norm": 0.6452786326408386,
"learning_rate": 0.00029335890253278475,
"loss": 1.9978923797607422,
"step": 7610
},
{
"epoch": 0.97536,
"grad_norm": 0.793510913848877,
"learning_rate": 0.0002933410972460036,
"loss": 1.9694734573364259,
"step": 7620
},
{
"epoch": 0.97664,
"grad_norm": 0.6846933364868164,
"learning_rate": 0.0002933232686643085,
"loss": 1.9795555114746093,
"step": 7630
},
{
"epoch": 0.97792,
"grad_norm": 0.6672417521476746,
"learning_rate": 0.00029330541679059666,
"loss": 2.008864974975586,
"step": 7640
},
{
"epoch": 0.9792,
"grad_norm": 0.6564120054244995,
"learning_rate": 0.0002932875416277694,
"loss": 1.9894664764404297,
"step": 7650
},
{
"epoch": 0.98048,
"grad_norm": 0.6869664788246155,
"learning_rate": 0.0002932696431787317,
"loss": 1.9923513412475586,
"step": 7660
},
{
"epoch": 0.98176,
"grad_norm": 0.601786732673645,
"learning_rate": 0.0002932517214463923,
"loss": 1.987978744506836,
"step": 7670
},
{
"epoch": 0.98304,
"grad_norm": 0.6519659757614136,
"learning_rate": 0.0002932337764336637,
"loss": 2.0119016647338865,
"step": 7680
},
{
"epoch": 0.98432,
"grad_norm": 0.7457683682441711,
"learning_rate": 0.00029321580814346213,
"loss": 2.004359817504883,
"step": 7690
},
{
"epoch": 0.9856,
"grad_norm": 0.6067475080490112,
"learning_rate": 0.00029319781657870784,
"loss": 1.9842674255371093,
"step": 7700
},
{
"epoch": 0.98688,
"grad_norm": 0.7065998315811157,
"learning_rate": 0.00029317980174232457,
"loss": 1.9875959396362304,
"step": 7710
},
{
"epoch": 0.98816,
"grad_norm": 0.6263285279273987,
"learning_rate": 0.0002931617636372401,
"loss": 2.007324981689453,
"step": 7720
},
{
"epoch": 0.98944,
"grad_norm": 0.6340311765670776,
"learning_rate": 0.00029314370226638577,
"loss": 1.9956018447875976,
"step": 7730
},
{
"epoch": 0.99072,
"grad_norm": 0.5806658267974854,
"learning_rate": 0.0002931256176326969,
"loss": 2.0026828765869142,
"step": 7740
},
{
"epoch": 0.992,
"grad_norm": 0.6932752132415771,
"learning_rate": 0.00029310750973911244,
"loss": 1.9967496871948243,
"step": 7750
},
{
"epoch": 0.99328,
"grad_norm": 0.5850876569747925,
"learning_rate": 0.00029308937858857516,
"loss": 1.9920431137084962,
"step": 7760
},
{
"epoch": 0.99456,
"grad_norm": 0.6746029853820801,
"learning_rate": 0.0002930712241840317,
"loss": 2.0010097503662108,
"step": 7770
},
{
"epoch": 0.99584,
"grad_norm": 0.6813861131668091,
"learning_rate": 0.00029305304652843224,
"loss": 1.9804662704467773,
"step": 7780
},
{
"epoch": 0.99712,
"grad_norm": 0.6215769052505493,
"learning_rate": 0.00029303484562473106,
"loss": 1.995717430114746,
"step": 7790
},
{
"epoch": 0.9984,
"grad_norm": 0.5907732248306274,
"learning_rate": 0.000293016621475886,
"loss": 2.0071685791015623,
"step": 7800
},
{
"epoch": 0.99968,
"grad_norm": 0.5828958749771118,
"learning_rate": 0.0002929983740848588,
"loss": 1.9976682662963867,
"step": 7810
},
{
"epoch": 1.0,
"eval_loss": 2.000473737716675,
"eval_runtime": 17.9727,
"eval_samples_per_second": 556.399,
"eval_steps_per_second": 69.55,
"step": 7813
}
],
"logging_steps": 10,
"max_steps": 78130,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1523666418204672.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}