gr00t_checkpoint / trainer_state.json
vlongle's picture
Upload folder using huggingface_hub
241d108 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.2,
"eval_steps": 500,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"grad_norm": 0.8453039526939392,
"learning_rate": 3.6e-06,
"loss": 1.2287,
"step": 10
},
{
"grad_norm": 0.22706136107444763,
"learning_rate": 7.6e-06,
"loss": 1.2158,
"step": 20
},
{
"grad_norm": 0.22833316028118134,
"learning_rate": 1.16e-05,
"loss": 1.2064,
"step": 30
},
{
"grad_norm": 0.2991141974925995,
"learning_rate": 1.56e-05,
"loss": 1.1477,
"step": 40
},
{
"grad_norm": 0.2874332666397095,
"learning_rate": 1.9600000000000002e-05,
"loss": 1.1045,
"step": 50
},
{
"grad_norm": 0.2624056935310364,
"learning_rate": 2.36e-05,
"loss": 1.0811,
"step": 60
},
{
"grad_norm": 0.2788711190223694,
"learning_rate": 2.7600000000000003e-05,
"loss": 1.057,
"step": 70
},
{
"grad_norm": 1.4561371803283691,
"learning_rate": 3.16e-05,
"loss": 0.9902,
"step": 80
},
{
"grad_norm": 0.8279716372489929,
"learning_rate": 3.56e-05,
"loss": 0.9047,
"step": 90
},
{
"grad_norm": 1.0444140434265137,
"learning_rate": 3.960000000000001e-05,
"loss": 0.8148,
"step": 100
},
{
"grad_norm": 1.9998027086257935,
"learning_rate": 4.36e-05,
"loss": 0.6477,
"step": 110
},
{
"grad_norm": 2.0924854278564453,
"learning_rate": 4.76e-05,
"loss": 0.4322,
"step": 120
},
{
"grad_norm": 1.3061543703079224,
"learning_rate": 5.16e-05,
"loss": 0.2847,
"step": 130
},
{
"grad_norm": 1.3199659585952759,
"learning_rate": 5.560000000000001e-05,
"loss": 0.2087,
"step": 140
},
{
"grad_norm": 0.7965296506881714,
"learning_rate": 5.96e-05,
"loss": 0.1827,
"step": 150
},
{
"grad_norm": 0.9607883095741272,
"learning_rate": 6.36e-05,
"loss": 0.1609,
"step": 160
},
{
"grad_norm": 0.8130837082862854,
"learning_rate": 6.76e-05,
"loss": 0.1486,
"step": 170
},
{
"grad_norm": 1.5720325708389282,
"learning_rate": 7.16e-05,
"loss": 0.1425,
"step": 180
},
{
"grad_norm": 1.010038137435913,
"learning_rate": 7.560000000000001e-05,
"loss": 0.1395,
"step": 190
},
{
"grad_norm": 1.1851860284805298,
"learning_rate": 7.960000000000001e-05,
"loss": 0.1346,
"step": 200
},
{
"grad_norm": 0.951931893825531,
"learning_rate": 8.36e-05,
"loss": 0.1284,
"step": 210
},
{
"grad_norm": 0.8772835731506348,
"learning_rate": 8.76e-05,
"loss": 0.1156,
"step": 220
},
{
"grad_norm": 0.7618266344070435,
"learning_rate": 9.16e-05,
"loss": 0.1112,
"step": 230
},
{
"grad_norm": 0.7764372229576111,
"learning_rate": 9.56e-05,
"loss": 0.112,
"step": 240
},
{
"grad_norm": 0.6424781680107117,
"learning_rate": 9.960000000000001e-05,
"loss": 0.1063,
"step": 250
},
{
"grad_norm": 0.6300824284553528,
"learning_rate": 0.00010360000000000001,
"loss": 0.106,
"step": 260
},
{
"grad_norm": 0.7829186916351318,
"learning_rate": 0.00010760000000000001,
"loss": 0.1049,
"step": 270
},
{
"grad_norm": 0.6146113276481628,
"learning_rate": 0.00011160000000000002,
"loss": 0.0996,
"step": 280
},
{
"grad_norm": 0.9386025071144104,
"learning_rate": 0.00011559999999999999,
"loss": 0.0994,
"step": 290
},
{
"grad_norm": 0.7355742454528809,
"learning_rate": 0.00011960000000000001,
"loss": 0.1023,
"step": 300
},
{
"grad_norm": 0.7597235441207886,
"learning_rate": 0.0001236,
"loss": 0.1022,
"step": 310
},
{
"grad_norm": 0.7152740955352783,
"learning_rate": 0.0001276,
"loss": 0.0997,
"step": 320
},
{
"grad_norm": 0.6420713067054749,
"learning_rate": 0.0001316,
"loss": 0.0942,
"step": 330
},
{
"grad_norm": 0.7762007117271423,
"learning_rate": 0.00013560000000000002,
"loss": 0.0961,
"step": 340
},
{
"grad_norm": 0.8098063468933105,
"learning_rate": 0.0001396,
"loss": 0.0972,
"step": 350
},
{
"grad_norm": 0.47828200459480286,
"learning_rate": 0.0001436,
"loss": 0.0972,
"step": 360
},
{
"grad_norm": 0.6147670149803162,
"learning_rate": 0.0001476,
"loss": 0.0921,
"step": 370
},
{
"grad_norm": 0.630636990070343,
"learning_rate": 0.0001516,
"loss": 0.0932,
"step": 380
},
{
"grad_norm": 0.7780483961105347,
"learning_rate": 0.00015560000000000001,
"loss": 0.0903,
"step": 390
},
{
"grad_norm": 0.549488365650177,
"learning_rate": 0.0001596,
"loss": 0.0911,
"step": 400
},
{
"grad_norm": 0.537593424320221,
"learning_rate": 0.0001636,
"loss": 0.0923,
"step": 410
},
{
"grad_norm": 0.35757097601890564,
"learning_rate": 0.0001676,
"loss": 0.0901,
"step": 420
},
{
"grad_norm": 0.5301028490066528,
"learning_rate": 0.0001716,
"loss": 0.0873,
"step": 430
},
{
"grad_norm": 0.5196937322616577,
"learning_rate": 0.0001756,
"loss": 0.0868,
"step": 440
},
{
"grad_norm": 0.5398963093757629,
"learning_rate": 0.0001796,
"loss": 0.0872,
"step": 450
},
{
"grad_norm": 0.39748314023017883,
"learning_rate": 0.00018360000000000002,
"loss": 0.0823,
"step": 460
},
{
"grad_norm": 0.47773298621177673,
"learning_rate": 0.0001876,
"loss": 0.0847,
"step": 470
},
{
"grad_norm": 0.6321383714675903,
"learning_rate": 0.0001916,
"loss": 0.0846,
"step": 480
},
{
"grad_norm": 0.43730923533439636,
"learning_rate": 0.0001956,
"loss": 0.0845,
"step": 490
},
{
"grad_norm": 0.3578919470310211,
"learning_rate": 0.0001996,
"loss": 0.0842,
"step": 500
},
{
"grad_norm": 0.5095338821411133,
"learning_rate": 0.000199999557098412,
"loss": 0.0819,
"step": 510
},
{
"grad_norm": 0.5493738651275635,
"learning_rate": 0.0001999980260856137,
"loss": 0.08,
"step": 520
},
{
"grad_norm": 0.42955926060676575,
"learning_rate": 0.00019999540151042328,
"loss": 0.0803,
"step": 530
},
{
"grad_norm": 0.42419669032096863,
"learning_rate": 0.0001999916834015426,
"loss": 0.0768,
"step": 540
},
{
"grad_norm": 0.3114933967590332,
"learning_rate": 0.0001999868717996323,
"loss": 0.0746,
"step": 550
},
{
"grad_norm": 0.4427943825721741,
"learning_rate": 0.00019998096675731135,
"loss": 0.0729,
"step": 560
},
{
"grad_norm": 0.2511034309864044,
"learning_rate": 0.0001999739683391563,
"loss": 0.0728,
"step": 570
},
{
"grad_norm": 0.30617302656173706,
"learning_rate": 0.00019996587662170074,
"loss": 0.0762,
"step": 580
},
{
"grad_norm": 0.3502658009529114,
"learning_rate": 0.00019995669169343438,
"loss": 0.0751,
"step": 590
},
{
"grad_norm": 0.3593088984489441,
"learning_rate": 0.00019994641365480214,
"loss": 0.0724,
"step": 600
},
{
"grad_norm": 0.3959539234638214,
"learning_rate": 0.00019993504261820298,
"loss": 0.0689,
"step": 610
},
{
"grad_norm": 0.40369659662246704,
"learning_rate": 0.0001999225787079888,
"loss": 0.0717,
"step": 620
},
{
"grad_norm": 0.25377678871154785,
"learning_rate": 0.00019990902206046287,
"loss": 0.07,
"step": 630
},
{
"grad_norm": 0.453866571187973,
"learning_rate": 0.00019989437282387856,
"loss": 0.0685,
"step": 640
},
{
"grad_norm": 0.4460633099079132,
"learning_rate": 0.00019987863115843748,
"loss": 0.0704,
"step": 650
},
{
"grad_norm": 0.35564717650413513,
"learning_rate": 0.00019986179723628804,
"loss": 0.0708,
"step": 660
},
{
"grad_norm": 0.256737619638443,
"learning_rate": 0.00019984387124152332,
"loss": 0.0682,
"step": 670
},
{
"grad_norm": 0.3440409302711487,
"learning_rate": 0.00019982485337017908,
"loss": 0.0672,
"step": 680
},
{
"grad_norm": 0.285584419965744,
"learning_rate": 0.00019980474383023174,
"loss": 0.0658,
"step": 690
},
{
"grad_norm": 0.2572714388370514,
"learning_rate": 0.00019978354284159604,
"loss": 0.065,
"step": 700
},
{
"grad_norm": 0.234798863530159,
"learning_rate": 0.00019976125063612252,
"loss": 0.0638,
"step": 710
},
{
"grad_norm": 0.30621594190597534,
"learning_rate": 0.00019973786745759525,
"loss": 0.0655,
"step": 720
},
{
"grad_norm": 0.3442394733428955,
"learning_rate": 0.00019971339356172885,
"loss": 0.0639,
"step": 730
},
{
"grad_norm": 0.3672217130661011,
"learning_rate": 0.00019968782921616596,
"loss": 0.0649,
"step": 740
},
{
"grad_norm": 0.2763133943080902,
"learning_rate": 0.00019966117470047418,
"loss": 0.0624,
"step": 750
},
{
"grad_norm": 0.3460654616355896,
"learning_rate": 0.000199633430306143,
"loss": 0.0614,
"step": 760
},
{
"grad_norm": 0.3609948456287384,
"learning_rate": 0.00019960459633658068,
"loss": 0.0612,
"step": 770
},
{
"grad_norm": 0.3908577263355255,
"learning_rate": 0.00019957467310711087,
"loss": 0.0596,
"step": 780
},
{
"grad_norm": 0.25239288806915283,
"learning_rate": 0.00019954366094496927,
"loss": 0.0612,
"step": 790
},
{
"grad_norm": 0.1670553982257843,
"learning_rate": 0.00019951156018929985,
"loss": 0.0645,
"step": 800
},
{
"grad_norm": 0.3140500485897064,
"learning_rate": 0.0001994783711911514,
"loss": 0.0582,
"step": 810
},
{
"grad_norm": 0.2637600898742676,
"learning_rate": 0.00019944409431347338,
"loss": 0.0608,
"step": 820
},
{
"grad_norm": 0.35186314582824707,
"learning_rate": 0.00019940872993111234,
"loss": 0.0602,
"step": 830
},
{
"grad_norm": 0.24156039953231812,
"learning_rate": 0.00019937227843080745,
"loss": 0.0597,
"step": 840
},
{
"grad_norm": 0.3593619465827942,
"learning_rate": 0.00019933474021118652,
"loss": 0.0582,
"step": 850
},
{
"grad_norm": 0.23397046327590942,
"learning_rate": 0.00019929611568276145,
"loss": 0.0583,
"step": 860
},
{
"grad_norm": 0.312755823135376,
"learning_rate": 0.0001992564052679239,
"loss": 0.0575,
"step": 870
},
{
"grad_norm": 0.3655824065208435,
"learning_rate": 0.00019921560940094066,
"loss": 0.0583,
"step": 880
},
{
"grad_norm": 0.46782079339027405,
"learning_rate": 0.00019917372852794874,
"loss": 0.0581,
"step": 890
},
{
"grad_norm": 0.25634610652923584,
"learning_rate": 0.00019913076310695068,
"loss": 0.0565,
"step": 900
},
{
"grad_norm": 0.4292739927768707,
"learning_rate": 0.0001990867136078094,
"loss": 0.054,
"step": 910
},
{
"grad_norm": 0.3035515546798706,
"learning_rate": 0.00019904158051224324,
"loss": 0.0564,
"step": 920
},
{
"grad_norm": 0.261796772480011,
"learning_rate": 0.00019899536431382045,
"loss": 0.0552,
"step": 930
},
{
"grad_norm": 0.24954535067081451,
"learning_rate": 0.000198948065517954,
"loss": 0.0551,
"step": 940
},
{
"grad_norm": 0.3103758990764618,
"learning_rate": 0.00019889968464189588,
"loss": 0.0558,
"step": 950
},
{
"grad_norm": 0.24562659859657288,
"learning_rate": 0.00019885022221473168,
"loss": 0.0555,
"step": 960
},
{
"grad_norm": 0.19912436604499817,
"learning_rate": 0.00019879967877737452,
"loss": 0.0528,
"step": 970
},
{
"grad_norm": 0.27990421652793884,
"learning_rate": 0.0001987480548825594,
"loss": 0.0519,
"step": 980
},
{
"grad_norm": 0.2675786018371582,
"learning_rate": 0.0001986953510948369,
"loss": 0.054,
"step": 990
},
{
"grad_norm": 0.26553845405578613,
"learning_rate": 0.00019864156799056723,
"loss": 0.0547,
"step": 1000
},
{
"grad_norm": 0.27253258228302,
"learning_rate": 0.00019858670615791377,
"loss": 0.0544,
"step": 1010
},
{
"grad_norm": 0.2688347399234772,
"learning_rate": 0.00019853076619683678,
"loss": 0.0518,
"step": 1020
},
{
"grad_norm": 0.2039746195077896,
"learning_rate": 0.00019847374871908668,
"loss": 0.0516,
"step": 1030
},
{
"grad_norm": 0.3064139783382416,
"learning_rate": 0.00019841565434819747,
"loss": 0.0506,
"step": 1040
},
{
"grad_norm": 0.24837756156921387,
"learning_rate": 0.00019835648371947987,
"loss": 0.0499,
"step": 1050
},
{
"grad_norm": 0.4394986629486084,
"learning_rate": 0.00019829623748001445,
"loss": 0.0502,
"step": 1060
},
{
"grad_norm": 0.1843259632587433,
"learning_rate": 0.00019823491628864436,
"loss": 0.0489,
"step": 1070
},
{
"grad_norm": 0.19856247305870056,
"learning_rate": 0.0001981725208159684,
"loss": 0.0482,
"step": 1080
},
{
"grad_norm": 0.24292084574699402,
"learning_rate": 0.0001981090517443334,
"loss": 0.0481,
"step": 1090
},
{
"grad_norm": 0.168592169880867,
"learning_rate": 0.000198044509767827,
"loss": 0.0488,
"step": 1100
},
{
"grad_norm": 0.21119219064712524,
"learning_rate": 0.0001979788955922699,
"loss": 0.0483,
"step": 1110
},
{
"grad_norm": 0.2026098072528839,
"learning_rate": 0.0001979122099352082,
"loss": 0.0506,
"step": 1120
},
{
"grad_norm": 0.3283786177635193,
"learning_rate": 0.0001978444535259056,
"loss": 0.0487,
"step": 1130
},
{
"grad_norm": 0.28264889121055603,
"learning_rate": 0.00019777562710533527,
"loss": 0.048,
"step": 1140
},
{
"grad_norm": 0.15788021683692932,
"learning_rate": 0.00019770573142617197,
"loss": 0.0485,
"step": 1150
},
{
"grad_norm": 0.12264436483383179,
"learning_rate": 0.00019763476725278363,
"loss": 0.0453,
"step": 1160
},
{
"grad_norm": 0.2884804606437683,
"learning_rate": 0.0001975627353612231,
"loss": 0.046,
"step": 1170
},
{
"grad_norm": 0.2817215919494629,
"learning_rate": 0.00019748963653921957,
"loss": 0.0455,
"step": 1180
},
{
"grad_norm": 0.27066612243652344,
"learning_rate": 0.00019741547158617006,
"loss": 0.0472,
"step": 1190
},
{
"grad_norm": 0.2489967942237854,
"learning_rate": 0.00019734024131313067,
"loss": 0.0448,
"step": 1200
},
{
"grad_norm": 0.3080730140209198,
"learning_rate": 0.00019726394654280752,
"loss": 0.0465,
"step": 1210
},
{
"grad_norm": 0.29908615350723267,
"learning_rate": 0.00019718658810954806,
"loss": 0.046,
"step": 1220
},
{
"grad_norm": 0.266740083694458,
"learning_rate": 0.0001971081668593317,
"loss": 0.0462,
"step": 1230
},
{
"grad_norm": 0.1910102367401123,
"learning_rate": 0.00019702868364976066,
"loss": 0.044,
"step": 1240
},
{
"grad_norm": 0.18283650279045105,
"learning_rate": 0.0001969481393500506,
"loss": 0.0456,
"step": 1250
},
{
"grad_norm": 0.2236376255750656,
"learning_rate": 0.0001968665348410211,
"loss": 0.0438,
"step": 1260
},
{
"grad_norm": 0.22641244530677795,
"learning_rate": 0.00019678387101508595,
"loss": 0.046,
"step": 1270
},
{
"grad_norm": 0.24328957498073578,
"learning_rate": 0.00019670014877624353,
"loss": 0.0449,
"step": 1280
},
{
"grad_norm": 0.13703007996082306,
"learning_rate": 0.00019661536904006682,
"loss": 0.044,
"step": 1290
},
{
"grad_norm": 0.23394407331943512,
"learning_rate": 0.00019652953273369342,
"loss": 0.0434,
"step": 1300
},
{
"grad_norm": 0.1836947351694107,
"learning_rate": 0.00019644264079581546,
"loss": 0.0426,
"step": 1310
},
{
"grad_norm": 0.2846074104309082,
"learning_rate": 0.00019635469417666923,
"loss": 0.0418,
"step": 1320
},
{
"grad_norm": 0.2773747742176056,
"learning_rate": 0.00019626569383802487,
"loss": 0.0469,
"step": 1330
},
{
"grad_norm": 0.21178005635738373,
"learning_rate": 0.00019617564075317584,
"loss": 0.0418,
"step": 1340
},
{
"grad_norm": 0.1820414513349533,
"learning_rate": 0.00019608453590692822,
"loss": 0.044,
"step": 1350
},
{
"grad_norm": 0.21951481699943542,
"learning_rate": 0.00019599238029559005,
"loss": 0.0422,
"step": 1360
},
{
"grad_norm": 0.24982789158821106,
"learning_rate": 0.00019589917492696026,
"loss": 0.043,
"step": 1370
},
{
"grad_norm": 0.1591809093952179,
"learning_rate": 0.00019580492082031792,
"loss": 0.042,
"step": 1380
},
{
"grad_norm": 0.23038744926452637,
"learning_rate": 0.00019570961900641076,
"loss": 0.0427,
"step": 1390
},
{
"grad_norm": 0.31900501251220703,
"learning_rate": 0.0001956132705274442,
"loss": 0.0415,
"step": 1400
},
{
"grad_norm": 0.3238053619861603,
"learning_rate": 0.00019551587643706976,
"loss": 0.0413,
"step": 1410
},
{
"grad_norm": 0.2005499303340912,
"learning_rate": 0.0001954174378003736,
"loss": 0.0412,
"step": 1420
},
{
"grad_norm": 0.19018618762493134,
"learning_rate": 0.00019531795569386487,
"loss": 0.0393,
"step": 1430
},
{
"grad_norm": 0.30942776799201965,
"learning_rate": 0.00019521743120546394,
"loss": 0.0419,
"step": 1440
},
{
"grad_norm": 0.2342258095741272,
"learning_rate": 0.00019511586543449053,
"loss": 0.0409,
"step": 1450
},
{
"grad_norm": 0.18405261635780334,
"learning_rate": 0.00019501325949165167,
"loss": 0.0407,
"step": 1460
},
{
"grad_norm": 0.2921549379825592,
"learning_rate": 0.00019490961449902946,
"loss": 0.041,
"step": 1470
},
{
"grad_norm": 0.18793050944805145,
"learning_rate": 0.00019480493159006895,
"loss": 0.0434,
"step": 1480
},
{
"grad_norm": 0.21637773513793945,
"learning_rate": 0.0001946992119095657,
"loss": 0.0407,
"step": 1490
},
{
"grad_norm": 0.2454444169998169,
"learning_rate": 0.00019459245661365313,
"loss": 0.0399,
"step": 1500
},
{
"grad_norm": 0.14031817018985748,
"learning_rate": 0.00019448466686979008,
"loss": 0.0374,
"step": 1510
},
{
"grad_norm": 0.26323285698890686,
"learning_rate": 0.00019437584385674793,
"loss": 0.0397,
"step": 1520
},
{
"grad_norm": 0.20086735486984253,
"learning_rate": 0.00019426598876459773,
"loss": 0.0395,
"step": 1530
},
{
"grad_norm": 0.2568920850753784,
"learning_rate": 0.0001941551027946971,
"loss": 0.0422,
"step": 1540
},
{
"grad_norm": 0.11995324492454529,
"learning_rate": 0.00019404318715967732,
"loss": 0.0398,
"step": 1550
},
{
"grad_norm": 0.17859072983264923,
"learning_rate": 0.00019393024308342983,
"loss": 0.039,
"step": 1560
},
{
"grad_norm": 0.2049851417541504,
"learning_rate": 0.0001938162718010929,
"loss": 0.0371,
"step": 1570
},
{
"grad_norm": 0.1736784279346466,
"learning_rate": 0.00019370127455903827,
"loss": 0.0393,
"step": 1580
},
{
"grad_norm": 0.16930192708969116,
"learning_rate": 0.0001935852526148573,
"loss": 0.0382,
"step": 1590
},
{
"grad_norm": 0.23014754056930542,
"learning_rate": 0.00019346820723734745,
"loss": 0.0401,
"step": 1600
},
{
"grad_norm": 0.27679961919784546,
"learning_rate": 0.00019335013970649817,
"loss": 0.0377,
"step": 1610
},
{
"grad_norm": 0.3353395462036133,
"learning_rate": 0.0001932310513134771,
"loss": 0.0411,
"step": 1620
},
{
"grad_norm": 0.2344261258840561,
"learning_rate": 0.00019311094336061578,
"loss": 0.0396,
"step": 1630
},
{
"grad_norm": 0.24594134092330933,
"learning_rate": 0.00019298981716139553,
"loss": 0.0395,
"step": 1640
},
{
"grad_norm": 0.16762520372867584,
"learning_rate": 0.00019286767404043316,
"loss": 0.0366,
"step": 1650
},
{
"grad_norm": 0.24015094339847565,
"learning_rate": 0.00019274451533346615,
"loss": 0.0358,
"step": 1660
},
{
"grad_norm": 0.1919541358947754,
"learning_rate": 0.00019262034238733843,
"loss": 0.0367,
"step": 1670
},
{
"grad_norm": 0.13164806365966797,
"learning_rate": 0.00019249515655998545,
"loss": 0.0348,
"step": 1680
},
{
"grad_norm": 0.23270496726036072,
"learning_rate": 0.00019236895922041943,
"loss": 0.0383,
"step": 1690
},
{
"grad_norm": 0.16496695578098297,
"learning_rate": 0.00019224175174871415,
"loss": 0.0383,
"step": 1700
},
{
"grad_norm": 0.17377837002277374,
"learning_rate": 0.00019211353553599017,
"loss": 0.0371,
"step": 1710
},
{
"grad_norm": 0.1682979315519333,
"learning_rate": 0.00019198431198439947,
"loss": 0.0383,
"step": 1720
},
{
"grad_norm": 0.2559478282928467,
"learning_rate": 0.0001918540825071101,
"loss": 0.0374,
"step": 1730
},
{
"grad_norm": 0.16148176789283752,
"learning_rate": 0.00019172284852829075,
"loss": 0.0346,
"step": 1740
},
{
"grad_norm": 0.21018460392951965,
"learning_rate": 0.0001915906114830952,
"loss": 0.0354,
"step": 1750
},
{
"grad_norm": 0.1702902913093567,
"learning_rate": 0.00019145737281764657,
"loss": 0.0365,
"step": 1760
},
{
"grad_norm": 0.1441521942615509,
"learning_rate": 0.0001913231339890216,
"loss": 0.0366,
"step": 1770
},
{
"grad_norm": 0.17645713686943054,
"learning_rate": 0.0001911878964652346,
"loss": 0.0365,
"step": 1780
},
{
"grad_norm": 0.2643643021583557,
"learning_rate": 0.00019105166172522138,
"loss": 0.0387,
"step": 1790
},
{
"grad_norm": 0.17437313497066498,
"learning_rate": 0.00019091443125882337,
"loss": 0.0391,
"step": 1800
},
{
"grad_norm": 0.17275704443454742,
"learning_rate": 0.00019077620656677085,
"loss": 0.036,
"step": 1810
},
{
"grad_norm": 0.17451389133930206,
"learning_rate": 0.00019063698916066697,
"loss": 0.0345,
"step": 1820
},
{
"grad_norm": 0.19119150936603546,
"learning_rate": 0.00019049678056297094,
"loss": 0.0352,
"step": 1830
},
{
"grad_norm": 0.17107735574245453,
"learning_rate": 0.00019035558230698154,
"loss": 0.035,
"step": 1840
},
{
"grad_norm": 0.19179688394069672,
"learning_rate": 0.00019021339593682028,
"loss": 0.0351,
"step": 1850
},
{
"grad_norm": 0.18375959992408752,
"learning_rate": 0.00019007022300741454,
"loss": 0.0374,
"step": 1860
},
{
"grad_norm": 0.242448091506958,
"learning_rate": 0.00018992606508448049,
"loss": 0.0356,
"step": 1870
},
{
"grad_norm": 0.20058102905750275,
"learning_rate": 0.00018978092374450611,
"loss": 0.0345,
"step": 1880
},
{
"grad_norm": 0.14428074657917023,
"learning_rate": 0.00018963480057473383,
"loss": 0.0343,
"step": 1890
},
{
"grad_norm": 0.18150261044502258,
"learning_rate": 0.00018948769717314328,
"loss": 0.0341,
"step": 1900
},
{
"grad_norm": 0.1219368577003479,
"learning_rate": 0.00018933961514843359,
"loss": 0.0327,
"step": 1910
},
{
"grad_norm": 0.24976667761802673,
"learning_rate": 0.0001891905561200061,
"loss": 0.0338,
"step": 1920
},
{
"grad_norm": 0.2597788870334625,
"learning_rate": 0.0001890405217179465,
"loss": 0.0349,
"step": 1930
},
{
"grad_norm": 0.18376731872558594,
"learning_rate": 0.00018888951358300696,
"loss": 0.0335,
"step": 1940
},
{
"grad_norm": 0.20351895689964294,
"learning_rate": 0.00018873753336658822,
"loss": 0.0344,
"step": 1950
},
{
"grad_norm": 0.2000509649515152,
"learning_rate": 0.00018858458273072158,
"loss": 0.0341,
"step": 1960
},
{
"grad_norm": 0.19135110080242157,
"learning_rate": 0.00018843066334805068,
"loss": 0.0322,
"step": 1970
},
{
"grad_norm": 0.1353762149810791,
"learning_rate": 0.00018827577690181332,
"loss": 0.0338,
"step": 1980
},
{
"grad_norm": 0.21709775924682617,
"learning_rate": 0.00018811992508582272,
"loss": 0.0349,
"step": 1990
},
{
"grad_norm": 0.21062973141670227,
"learning_rate": 0.0001879631096044495,
"loss": 0.0327,
"step": 2000
}
],
"logging_steps": 10,
"max_steps": 10000,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 320,
"trial_name": null,
"trial_params": null
}