toy_model_2 / trainer_state.json
VladShash's picture
Upload 14 files
850113f verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1582,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006324110671936759,
"grad_norm": 9.46192455291748,
"learning_rate": 9.999201453419822e-06,
"loss": 1.0985,
"step": 10
},
{
"epoch": 0.012648221343873518,
"grad_norm": 11.03305435180664,
"learning_rate": 9.99644137295895e-06,
"loss": 0.8008,
"step": 20
},
{
"epoch": 0.018972332015810278,
"grad_norm": 8.343941688537598,
"learning_rate": 9.991710988172794e-06,
"loss": 0.7198,
"step": 30
},
{
"epoch": 0.025296442687747035,
"grad_norm": 9.01296329498291,
"learning_rate": 9.98501216444851e-06,
"loss": 0.6853,
"step": 40
},
{
"epoch": 0.03162055335968379,
"grad_norm": 7.772639274597168,
"learning_rate": 9.976347543410487e-06,
"loss": 0.5944,
"step": 50
},
{
"epoch": 0.037944664031620556,
"grad_norm": 8.78039264678955,
"learning_rate": 9.965720541878623e-06,
"loss": 0.5808,
"step": 60
},
{
"epoch": 0.04426877470355731,
"grad_norm": 11.329726219177246,
"learning_rate": 9.953135350520935e-06,
"loss": 0.4663,
"step": 70
},
{
"epoch": 0.05059288537549407,
"grad_norm": 7.309874057769775,
"learning_rate": 9.938596932201018e-06,
"loss": 0.4968,
"step": 80
},
{
"epoch": 0.05691699604743083,
"grad_norm": 9.987103462219238,
"learning_rate": 9.922111020020963e-06,
"loss": 0.4906,
"step": 90
},
{
"epoch": 0.06324110671936758,
"grad_norm": 7.361323356628418,
"learning_rate": 9.90368411506057e-06,
"loss": 0.4997,
"step": 100
},
{
"epoch": 0.06956521739130435,
"grad_norm": 6.6940693855285645,
"learning_rate": 9.883323483813697e-06,
"loss": 0.3826,
"step": 110
},
{
"epoch": 0.07588932806324111,
"grad_norm": 5.134791851043701,
"learning_rate": 9.861037155322777e-06,
"loss": 0.3218,
"step": 120
},
{
"epoch": 0.08221343873517786,
"grad_norm": 6.337493896484375,
"learning_rate": 9.836833918012642e-06,
"loss": 0.3838,
"step": 130
},
{
"epoch": 0.08853754940711463,
"grad_norm": 9.632590293884277,
"learning_rate": 9.810723316224872e-06,
"loss": 0.3447,
"step": 140
},
{
"epoch": 0.09486166007905138,
"grad_norm": 9.018434524536133,
"learning_rate": 9.782715646454073e-06,
"loss": 0.3119,
"step": 150
},
{
"epoch": 0.10118577075098814,
"grad_norm": 8.715914726257324,
"learning_rate": 9.752821953287537e-06,
"loss": 0.2992,
"step": 160
},
{
"epoch": 0.1075098814229249,
"grad_norm": 7.031384468078613,
"learning_rate": 9.721054025049902e-06,
"loss": 0.4097,
"step": 170
},
{
"epoch": 0.11383399209486166,
"grad_norm": 6.2494797706604,
"learning_rate": 9.687424389154528e-06,
"loss": 0.3309,
"step": 180
},
{
"epoch": 0.12015810276679842,
"grad_norm": 5.795834541320801,
"learning_rate": 9.651946307163417e-06,
"loss": 0.3737,
"step": 190
},
{
"epoch": 0.12648221343873517,
"grad_norm": 5.849041938781738,
"learning_rate": 9.614633769557635e-06,
"loss": 0.2506,
"step": 200
},
{
"epoch": 0.13280632411067195,
"grad_norm": 4.081539154052734,
"learning_rate": 9.57550149022027e-06,
"loss": 0.2744,
"step": 210
},
{
"epoch": 0.1391304347826087,
"grad_norm": 6.849961757659912,
"learning_rate": 9.534564900634146e-06,
"loss": 0.2722,
"step": 220
},
{
"epoch": 0.14545454545454545,
"grad_norm": 6.057841777801514,
"learning_rate": 9.491840143796554e-06,
"loss": 0.2908,
"step": 230
},
{
"epoch": 0.15177865612648223,
"grad_norm": 6.073554039001465,
"learning_rate": 9.44734406785338e-06,
"loss": 0.3071,
"step": 240
},
{
"epoch": 0.15810276679841898,
"grad_norm": 3.27799391746521,
"learning_rate": 9.401094219455202e-06,
"loss": 0.2752,
"step": 250
},
{
"epoch": 0.16442687747035573,
"grad_norm": 7.029287815093994,
"learning_rate": 9.353108836837907e-06,
"loss": 0.2446,
"step": 260
},
{
"epoch": 0.1707509881422925,
"grad_norm": 3.119824171066284,
"learning_rate": 9.30340684263059e-06,
"loss": 0.2597,
"step": 270
},
{
"epoch": 0.17707509881422925,
"grad_norm": 4.375607490539551,
"learning_rate": 9.252007836393593e-06,
"loss": 0.2314,
"step": 280
},
{
"epoch": 0.183399209486166,
"grad_norm": 9.691184043884277,
"learning_rate": 9.198932086889567e-06,
"loss": 0.2789,
"step": 290
},
{
"epoch": 0.18972332015810275,
"grad_norm": 10.60154914855957,
"learning_rate": 9.14420052409067e-06,
"loss": 0.258,
"step": 300
},
{
"epoch": 0.19604743083003953,
"grad_norm": 3.9770102500915527,
"learning_rate": 9.087834730925002e-06,
"loss": 0.2302,
"step": 310
},
{
"epoch": 0.20237154150197628,
"grad_norm": 4.464337348937988,
"learning_rate": 9.029856934765565e-06,
"loss": 0.1974,
"step": 320
},
{
"epoch": 0.20869565217391303,
"grad_norm": 9.647624015808105,
"learning_rate": 8.970289998665083e-06,
"loss": 0.2763,
"step": 330
},
{
"epoch": 0.2150197628458498,
"grad_norm": 4.888010025024414,
"learning_rate": 8.90915741234015e-06,
"loss": 0.2278,
"step": 340
},
{
"epoch": 0.22134387351778656,
"grad_norm": 5.202849388122559,
"learning_rate": 8.846483282908267e-06,
"loss": 0.2957,
"step": 350
},
{
"epoch": 0.2276679841897233,
"grad_norm": 6.026373863220215,
"learning_rate": 8.782292325381397e-06,
"loss": 0.2376,
"step": 360
},
{
"epoch": 0.2339920948616601,
"grad_norm": 4.311495304107666,
"learning_rate": 8.716609852919808e-06,
"loss": 0.2793,
"step": 370
},
{
"epoch": 0.24031620553359684,
"grad_norm": 6.434778213500977,
"learning_rate": 8.649461766850053e-06,
"loss": 0.1864,
"step": 380
},
{
"epoch": 0.2466403162055336,
"grad_norm": 6.526483535766602,
"learning_rate": 8.580874546450992e-06,
"loss": 0.2317,
"step": 390
},
{
"epoch": 0.25296442687747034,
"grad_norm": 4.416475772857666,
"learning_rate": 8.510875238511911e-06,
"loss": 0.207,
"step": 400
},
{
"epoch": 0.2592885375494071,
"grad_norm": 5.223466396331787,
"learning_rate": 8.439491446666858e-06,
"loss": 0.2325,
"step": 410
},
{
"epoch": 0.2656126482213439,
"grad_norm": 7.620017051696777,
"learning_rate": 8.366751320509383e-06,
"loss": 0.2968,
"step": 420
},
{
"epoch": 0.2719367588932806,
"grad_norm": 4.467705726623535,
"learning_rate": 8.292683544491974e-06,
"loss": 0.2003,
"step": 430
},
{
"epoch": 0.2782608695652174,
"grad_norm": 8.371648788452148,
"learning_rate": 8.217317326614598e-06,
"loss": 0.1876,
"step": 440
},
{
"epoch": 0.2845849802371542,
"grad_norm": 4.598704814910889,
"learning_rate": 8.14068238690677e-06,
"loss": 0.236,
"step": 450
},
{
"epoch": 0.2909090909090909,
"grad_norm": 7.2356181144714355,
"learning_rate": 8.062808945707705e-06,
"loss": 0.2488,
"step": 460
},
{
"epoch": 0.2972332015810277,
"grad_norm": 7.233348846435547,
"learning_rate": 7.983727711749194e-06,
"loss": 0.2399,
"step": 470
},
{
"epoch": 0.30355731225296445,
"grad_norm": 3.6903178691864014,
"learning_rate": 7.903469870045869e-06,
"loss": 0.2155,
"step": 480
},
{
"epoch": 0.3098814229249012,
"grad_norm": 2.80163311958313,
"learning_rate": 7.82206706959765e-06,
"loss": 0.1961,
"step": 490
},
{
"epoch": 0.31620553359683795,
"grad_norm": 5.6080241203308105,
"learning_rate": 7.739551410909254e-06,
"loss": 0.2785,
"step": 500
},
{
"epoch": 0.32252964426877473,
"grad_norm": 5.316874980926514,
"learning_rate": 7.6559554333316e-06,
"loss": 0.2273,
"step": 510
},
{
"epoch": 0.32885375494071145,
"grad_norm": 6.32692289352417,
"learning_rate": 7.571312102230218e-06,
"loss": 0.1419,
"step": 520
},
{
"epoch": 0.33517786561264823,
"grad_norm": 17.47705078125,
"learning_rate": 7.485654795985626e-06,
"loss": 0.3048,
"step": 530
},
{
"epoch": 0.341501976284585,
"grad_norm": 2.7128899097442627,
"learning_rate": 7.399017292830848e-06,
"loss": 0.1701,
"step": 540
},
{
"epoch": 0.34782608695652173,
"grad_norm": 10.415395736694336,
"learning_rate": 7.311433757531252e-06,
"loss": 0.2112,
"step": 550
},
{
"epoch": 0.3541501976284585,
"grad_norm": 5.806975841522217,
"learning_rate": 7.222938727911974e-06,
"loss": 0.2398,
"step": 560
},
{
"epoch": 0.36047430830039523,
"grad_norm": 5.774322986602783,
"learning_rate": 7.133567101238209e-06,
"loss": 0.174,
"step": 570
},
{
"epoch": 0.366798418972332,
"grad_norm": 3.898362159729004,
"learning_rate": 7.043354120453771e-06,
"loss": 0.1869,
"step": 580
},
{
"epoch": 0.3731225296442688,
"grad_norm": 4.964472770690918,
"learning_rate": 6.9523353602833485e-06,
"loss": 0.186,
"step": 590
},
{
"epoch": 0.3794466403162055,
"grad_norm": 7.7693190574646,
"learning_rate": 6.860546713203895e-06,
"loss": 0.2266,
"step": 600
},
{
"epoch": 0.3857707509881423,
"grad_norm": 6.536276817321777,
"learning_rate": 6.768024375290747e-06,
"loss": 0.2001,
"step": 610
},
{
"epoch": 0.39209486166007906,
"grad_norm": 13.879557609558105,
"learning_rate": 6.6748048319440075e-06,
"loss": 0.2414,
"step": 620
},
{
"epoch": 0.3984189723320158,
"grad_norm": 3.777066230773926,
"learning_rate": 6.580924843500833e-06,
"loss": 0.2269,
"step": 630
},
{
"epoch": 0.40474308300395256,
"grad_norm": 2.5478343963623047,
"learning_rate": 6.486421430739309e-06,
"loss": 0.1793,
"step": 640
},
{
"epoch": 0.41106719367588934,
"grad_norm": 3.3290770053863525,
"learning_rate": 6.391331860279623e-06,
"loss": 0.2254,
"step": 650
},
{
"epoch": 0.41739130434782606,
"grad_norm": 3.926723003387451,
"learning_rate": 6.295693629888277e-06,
"loss": 0.2365,
"step": 660
},
{
"epoch": 0.42371541501976284,
"grad_norm": 2.598579168319702,
"learning_rate": 6.199544453691175e-06,
"loss": 0.2057,
"step": 670
},
{
"epoch": 0.4300395256916996,
"grad_norm": 3.4086475372314453,
"learning_rate": 6.1029222473013705e-06,
"loss": 0.2235,
"step": 680
},
{
"epoch": 0.43636363636363634,
"grad_norm": 7.923166275024414,
"learning_rate": 6.005865112867362e-06,
"loss": 0.2206,
"step": 690
},
{
"epoch": 0.4426877470355731,
"grad_norm": 4.715216636657715,
"learning_rate": 5.908411324047844e-06,
"loss": 0.1337,
"step": 700
},
{
"epoch": 0.4490118577075099,
"grad_norm": 2.680840253829956,
"learning_rate": 5.810599310918796e-06,
"loss": 0.2243,
"step": 710
},
{
"epoch": 0.4553359683794466,
"grad_norm": 6.077740669250488,
"learning_rate": 5.712467644818935e-06,
"loss": 0.2135,
"step": 720
},
{
"epoch": 0.4616600790513834,
"grad_norm": 5.459410190582275,
"learning_rate": 5.614055023139404e-06,
"loss": 0.2081,
"step": 730
},
{
"epoch": 0.4679841897233202,
"grad_norm": 3.2688136100769043,
"learning_rate": 5.5154002540638096e-06,
"loss": 0.2239,
"step": 740
},
{
"epoch": 0.4743083003952569,
"grad_norm": 4.673855304718018,
"learning_rate": 5.416542241264524e-06,
"loss": 0.1613,
"step": 750
},
{
"epoch": 0.4806324110671937,
"grad_norm": 5.865203857421875,
"learning_rate": 5.317519968561367e-06,
"loss": 0.2067,
"step": 760
},
{
"epoch": 0.48695652173913045,
"grad_norm": 8.180610656738281,
"learning_rate": 5.2183724845486525e-06,
"loss": 0.2231,
"step": 770
},
{
"epoch": 0.4932806324110672,
"grad_norm": 4.828525066375732,
"learning_rate": 5.119138887196724e-06,
"loss": 0.1318,
"step": 780
},
{
"epoch": 0.49960474308300395,
"grad_norm": 2.936441659927368,
"learning_rate": 5.019858308433979e-06,
"loss": 0.1463,
"step": 790
},
{
"epoch": 0.5059288537549407,
"grad_norm": 3.418560743331909,
"learning_rate": 4.920569898715544e-06,
"loss": 0.1743,
"step": 800
},
{
"epoch": 0.5122529644268775,
"grad_norm": 4.025258541107178,
"learning_rate": 4.8213128115846136e-06,
"loss": 0.1409,
"step": 810
},
{
"epoch": 0.5185770750988142,
"grad_norm": 3.7686188220977783,
"learning_rate": 4.722126188232586e-06,
"loss": 0.1306,
"step": 820
},
{
"epoch": 0.524901185770751,
"grad_norm": 5.471837043762207,
"learning_rate": 4.623049142064066e-06,
"loss": 0.16,
"step": 830
},
{
"epoch": 0.5312252964426878,
"grad_norm": 4.224705696105957,
"learning_rate": 4.524120743272827e-06,
"loss": 0.2751,
"step": 840
},
{
"epoch": 0.5375494071146245,
"grad_norm": 2.844553232192993,
"learning_rate": 4.4253800034348e-06,
"loss": 0.1805,
"step": 850
},
{
"epoch": 0.5438735177865612,
"grad_norm": 3.4316790103912354,
"learning_rate": 4.326865860124203e-06,
"loss": 0.2133,
"step": 860
},
{
"epoch": 0.5501976284584981,
"grad_norm": 8.068376541137695,
"learning_rate": 4.228617161558837e-06,
"loss": 0.1663,
"step": 870
},
{
"epoch": 0.5565217391304348,
"grad_norm": 4.661905288696289,
"learning_rate": 4.130672651280612e-06,
"loss": 0.1606,
"step": 880
},
{
"epoch": 0.5628458498023715,
"grad_norm": 4.84809684753418,
"learning_rate": 4.033070952877362e-06,
"loss": 0.163,
"step": 890
},
{
"epoch": 0.5691699604743083,
"grad_norm": 2.1477699279785156,
"learning_rate": 3.935850554751964e-06,
"loss": 0.1557,
"step": 900
},
{
"epoch": 0.5754940711462451,
"grad_norm": 2.225597620010376,
"learning_rate": 3.839049794944767e-06,
"loss": 0.1597,
"step": 910
},
{
"epoch": 0.5818181818181818,
"grad_norm": 3.4657671451568604,
"learning_rate": 3.7427068460152977e-06,
"loss": 0.1516,
"step": 920
},
{
"epoch": 0.5881422924901186,
"grad_norm": 5.109956741333008,
"learning_rate": 3.6468596999892415e-06,
"loss": 0.2049,
"step": 930
},
{
"epoch": 0.5944664031620553,
"grad_norm": 7.530884742736816,
"learning_rate": 3.551546153376611e-06,
"loss": 0.1816,
"step": 940
},
{
"epoch": 0.6007905138339921,
"grad_norm": 3.20400071144104,
"learning_rate": 3.4568037922670135e-06,
"loss": 0.2121,
"step": 950
},
{
"epoch": 0.6071146245059289,
"grad_norm": 2.265789270401001,
"learning_rate": 3.3626699775078884e-06,
"loss": 0.1419,
"step": 960
},
{
"epoch": 0.6134387351778656,
"grad_norm": 5.862637996673584,
"learning_rate": 3.269181829971581e-06,
"loss": 0.2162,
"step": 970
},
{
"epoch": 0.6197628458498023,
"grad_norm": 3.612271547317505,
"learning_rate": 3.1763762159170317e-06,
"loss": 0.1425,
"step": 980
},
{
"epoch": 0.6260869565217392,
"grad_norm": 6.073237895965576,
"learning_rate": 3.08428973245189e-06,
"loss": 0.2775,
"step": 990
},
{
"epoch": 0.6324110671936759,
"grad_norm": 4.313070297241211,
"learning_rate": 2.9929586931007494e-06,
"loss": 0.1259,
"step": 1000
},
{
"epoch": 0.6387351778656126,
"grad_norm": 3.83176326751709,
"learning_rate": 2.9024191134852153e-06,
"loss": 0.1277,
"step": 1010
},
{
"epoch": 0.6450592885375495,
"grad_norm": 5.826846122741699,
"learning_rate": 2.8127066971214533e-06,
"loss": 0.1958,
"step": 1020
},
{
"epoch": 0.6513833992094862,
"grad_norm": 4.473873138427734,
"learning_rate": 2.723856821340806e-06,
"loss": 0.1839,
"step": 1030
},
{
"epoch": 0.6577075098814229,
"grad_norm": 4.097567081451416,
"learning_rate": 2.6359045233390474e-06,
"loss": 0.1879,
"step": 1040
},
{
"epoch": 0.6640316205533597,
"grad_norm": 2.163008689880371,
"learning_rate": 2.5488844863597613e-06,
"loss": 0.1452,
"step": 1050
},
{
"epoch": 0.6703557312252965,
"grad_norm": 1.4688968658447266,
"learning_rate": 2.4628310260172838e-06,
"loss": 0.1601,
"step": 1060
},
{
"epoch": 0.6766798418972332,
"grad_norm": 1.575208067893982,
"learning_rate": 2.3777780767646556e-06,
"loss": 0.1346,
"step": 1070
},
{
"epoch": 0.68300395256917,
"grad_norm": 1.3976165056228638,
"learning_rate": 2.2937591785118407e-06,
"loss": 0.2126,
"step": 1080
},
{
"epoch": 0.6893280632411067,
"grad_norm": 4.7377543449401855,
"learning_rate": 2.210807463399554e-06,
"loss": 0.148,
"step": 1090
},
{
"epoch": 0.6956521739130435,
"grad_norm": 2.857984781265259,
"learning_rate": 2.128955642733877e-06,
"loss": 0.2124,
"step": 1100
},
{
"epoch": 0.7019762845849803,
"grad_norm": 2.6081042289733887,
"learning_rate": 2.0482359940868336e-06,
"loss": 0.1758,
"step": 1110
},
{
"epoch": 0.708300395256917,
"grad_norm": 4.629555702209473,
"learning_rate": 1.9686803485680095e-06,
"loss": 0.1821,
"step": 1120
},
{
"epoch": 0.7146245059288537,
"grad_norm": 1.8115971088409424,
"learning_rate": 1.890320078272212e-06,
"loss": 0.2113,
"step": 1130
},
{
"epoch": 0.7209486166007905,
"grad_norm": 3.0228452682495117,
"learning_rate": 1.8131860839081634e-06,
"loss": 0.1316,
"step": 1140
},
{
"epoch": 0.7272727272727273,
"grad_norm": 2.8287572860717773,
"learning_rate": 1.737308782613053e-06,
"loss": 0.1586,
"step": 1150
},
{
"epoch": 0.733596837944664,
"grad_norm": 17.451133728027344,
"learning_rate": 1.6627180959578143e-06,
"loss": 0.1983,
"step": 1160
},
{
"epoch": 0.7399209486166007,
"grad_norm": 2.922147274017334,
"learning_rate": 1.589443438147789e-06,
"loss": 0.1311,
"step": 1170
},
{
"epoch": 0.7462450592885376,
"grad_norm": 3.4132003784179688,
"learning_rate": 1.517513704423501e-06,
"loss": 0.1622,
"step": 1180
},
{
"epoch": 0.7525691699604743,
"grad_norm": 6.654974460601807,
"learning_rate": 1.4469572596660486e-06,
"loss": 0.1876,
"step": 1190
},
{
"epoch": 0.758893280632411,
"grad_norm": 2.83597731590271,
"learning_rate": 1.377801927211666e-06,
"loss": 0.1505,
"step": 1200
},
{
"epoch": 0.7652173913043478,
"grad_norm": 2.1424484252929688,
"learning_rate": 1.3100749778798166e-06,
"loss": 0.146,
"step": 1210
},
{
"epoch": 0.7715415019762846,
"grad_norm": 5.101158142089844,
"learning_rate": 1.2438031192191857e-06,
"loss": 0.1553,
"step": 1220
},
{
"epoch": 0.7778656126482213,
"grad_norm": 6.358023166656494,
"learning_rate": 1.1790124849757733e-06,
"loss": 0.1563,
"step": 1230
},
{
"epoch": 0.7841897233201581,
"grad_norm": 7.944547176361084,
"learning_rate": 1.1157286247872873e-06,
"loss": 0.1525,
"step": 1240
},
{
"epoch": 0.7905138339920948,
"grad_norm": 5.659843444824219,
"learning_rate": 1.0539764941078451e-06,
"loss": 0.1853,
"step": 1250
},
{
"epoch": 0.7968379446640316,
"grad_norm": 2.0195701122283936,
"learning_rate": 9.937804443670136e-07,
"loss": 0.167,
"step": 1260
},
{
"epoch": 0.8031620553359684,
"grad_norm": 3.8951923847198486,
"learning_rate": 9.351642133670258e-07,
"loss": 0.1456,
"step": 1270
},
{
"epoch": 0.8094861660079051,
"grad_norm": 2.7416598796844482,
"learning_rate": 8.781509159219748e-07,
"loss": 0.1427,
"step": 1280
},
{
"epoch": 0.8158102766798419,
"grad_norm": 5.3337812423706055,
"learning_rate": 8.227630347426868e-07,
"loss": 0.1791,
"step": 1290
},
{
"epoch": 0.8221343873517787,
"grad_norm": 4.012369155883789,
"learning_rate": 7.690224115708573e-07,
"loss": 0.1287,
"step": 1300
},
{
"epoch": 0.8284584980237154,
"grad_norm": 2.7309205532073975,
"learning_rate": 7.16950238565941e-07,
"loss": 0.1168,
"step": 1310
},
{
"epoch": 0.8347826086956521,
"grad_norm": 2.533595561981201,
"learning_rate": 6.66567049948203e-07,
"loss": 0.1398,
"step": 1320
},
{
"epoch": 0.841106719367589,
"grad_norm": 5.537268161773682,
"learning_rate": 6.178927139012242e-07,
"loss": 0.1368,
"step": 1330
},
{
"epoch": 0.8474308300395257,
"grad_norm": 3.948707103729248,
"learning_rate": 5.709464247370545e-07,
"loss": 0.1866,
"step": 1340
},
{
"epoch": 0.8537549407114624,
"grad_norm": 2.8135249614715576,
"learning_rate": 5.257466953270951e-07,
"loss": 0.1653,
"step": 1350
},
{
"epoch": 0.8600790513833992,
"grad_norm": 16.620941162109375,
"learning_rate": 4.823113498017024e-07,
"loss": 0.1467,
"step": 1360
},
{
"epoch": 0.866403162055336,
"grad_norm": 2.369086980819702,
"learning_rate": 4.406575165213972e-07,
"loss": 0.1163,
"step": 1370
},
{
"epoch": 0.8727272727272727,
"grad_norm": 8.848119735717773,
"learning_rate": 4.008016213224408e-07,
"loss": 0.2006,
"step": 1380
},
{
"epoch": 0.8790513833992095,
"grad_norm": 3.858823537826538,
"learning_rate": 3.6275938103944477e-07,
"loss": 0.1547,
"step": 1390
},
{
"epoch": 0.8853754940711462,
"grad_norm": 2.8297841548919678,
"learning_rate": 3.2654579730756695e-07,
"loss": 0.1401,
"step": 1400
},
{
"epoch": 0.891699604743083,
"grad_norm": 1.4849733114242554,
"learning_rate": 2.921751506467479e-07,
"loss": 0.1475,
"step": 1410
},
{
"epoch": 0.8980237154150198,
"grad_norm": 3.9321346282958984,
"learning_rate": 2.596609948303097e-07,
"loss": 0.1242,
"step": 1420
},
{
"epoch": 0.9043478260869565,
"grad_norm": 4.110458850860596,
"learning_rate": 2.290161515401368e-07,
"loss": 0.1948,
"step": 1430
},
{
"epoch": 0.9106719367588932,
"grad_norm": 6.044195652008057,
"learning_rate": 2.0025270531055518e-07,
"loss": 0.2341,
"step": 1440
},
{
"epoch": 0.9169960474308301,
"grad_norm": 2.854436159133911,
"learning_rate": 1.7338199876289984e-07,
"loss": 0.1566,
"step": 1450
},
{
"epoch": 0.9233201581027668,
"grad_norm": 2.5922341346740723,
"learning_rate": 1.4841462813264763e-07,
"loss": 0.1149,
"step": 1460
},
{
"epoch": 0.9296442687747035,
"grad_norm": 5.463244438171387,
"learning_rate": 1.253604390908819e-07,
"loss": 0.186,
"step": 1470
},
{
"epoch": 0.9359683794466404,
"grad_norm": 13.760538101196289,
"learning_rate": 1.0422852286173512e-07,
"loss": 0.181,
"step": 1480
},
{
"epoch": 0.9422924901185771,
"grad_norm": 3.221524715423584,
"learning_rate": 8.502721263734082e-08,
"loss": 0.1553,
"step": 1490
},
{
"epoch": 0.9486166007905138,
"grad_norm": 3.175731658935547,
"learning_rate": 6.776408029171178e-08,
"loss": 0.1141,
"step": 1500
},
{
"epoch": 0.9549407114624506,
"grad_norm": 5.117398262023926,
"learning_rate": 5.244593339483406e-08,
"loss": 0.1096,
"step": 1510
},
{
"epoch": 0.9612648221343874,
"grad_norm": 10.858709335327148,
"learning_rate": 3.907881252816048e-08,
"loss": 0.16,
"step": 1520
},
{
"epoch": 0.9675889328063241,
"grad_norm": 4.956307888031006,
"learning_rate": 2.7667988902557174e-08,
"loss": 0.1695,
"step": 1530
},
{
"epoch": 0.9739130434782609,
"grad_norm": 2.5706870555877686,
"learning_rate": 1.821796227964856e-08,
"loss": 0.1398,
"step": 1540
},
{
"epoch": 0.9802371541501976,
"grad_norm": 4.848339557647705,
"learning_rate": 1.0732459197371758e-08,
"loss": 0.1581,
"step": 1550
},
{
"epoch": 0.9865612648221344,
"grad_norm": 6.434995174407959,
"learning_rate": 5.214431500449868e-09,
"loss": 0.1569,
"step": 1560
},
{
"epoch": 0.9928853754940712,
"grad_norm": 5.2261962890625,
"learning_rate": 1.6660551763564292e-09,
"loss": 0.1485,
"step": 1570
},
{
"epoch": 0.9992094861660079,
"grad_norm": 2.2790348529815674,
"learning_rate": 8.87294972318209e-11,
"loss": 0.1657,
"step": 1580
}
],
"logging_steps": 10,
"max_steps": 1582,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.4231129196199936e+16,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}