toy_model_3 / trainer_state.json
VladShash's picture
Upload 15 files
b41e77e 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": 6.975515365600586,
"learning_rate": 9.999201453419822e-06,
"loss": 0.949,
"step": 10
},
{
"epoch": 0.012648221343873518,
"grad_norm": 9.093941688537598,
"learning_rate": 9.99644137295895e-06,
"loss": 0.6659,
"step": 20
},
{
"epoch": 0.018972332015810278,
"grad_norm": 7.245123386383057,
"learning_rate": 9.991710988172794e-06,
"loss": 0.6138,
"step": 30
},
{
"epoch": 0.025296442687747035,
"grad_norm": 8.16720962524414,
"learning_rate": 9.98501216444851e-06,
"loss": 0.6027,
"step": 40
},
{
"epoch": 0.03162055335968379,
"grad_norm": 6.609710216522217,
"learning_rate": 9.976347543410487e-06,
"loss": 0.5173,
"step": 50
},
{
"epoch": 0.037944664031620556,
"grad_norm": 7.609875202178955,
"learning_rate": 9.965720541878623e-06,
"loss": 0.5046,
"step": 60
},
{
"epoch": 0.04426877470355731,
"grad_norm": 10.813131332397461,
"learning_rate": 9.953135350520935e-06,
"loss": 0.3816,
"step": 70
},
{
"epoch": 0.05059288537549407,
"grad_norm": 6.553898811340332,
"learning_rate": 9.938596932201018e-06,
"loss": 0.4429,
"step": 80
},
{
"epoch": 0.05691699604743083,
"grad_norm": 7.274947166442871,
"learning_rate": 9.922111020020963e-06,
"loss": 0.4449,
"step": 90
},
{
"epoch": 0.06324110671936758,
"grad_norm": 6.7943339347839355,
"learning_rate": 9.90368411506057e-06,
"loss": 0.4448,
"step": 100
},
{
"epoch": 0.06956521739130435,
"grad_norm": 5.958676338195801,
"learning_rate": 9.883323483813697e-06,
"loss": 0.3659,
"step": 110
},
{
"epoch": 0.07588932806324111,
"grad_norm": 4.650176525115967,
"learning_rate": 9.861037155322777e-06,
"loss": 0.3042,
"step": 120
},
{
"epoch": 0.08221343873517786,
"grad_norm": 6.156113147735596,
"learning_rate": 9.836833918012642e-06,
"loss": 0.3697,
"step": 130
},
{
"epoch": 0.08853754940711463,
"grad_norm": 9.764841079711914,
"learning_rate": 9.810723316224872e-06,
"loss": 0.3407,
"step": 140
},
{
"epoch": 0.09486166007905138,
"grad_norm": 6.659021377563477,
"learning_rate": 9.782715646454073e-06,
"loss": 0.311,
"step": 150
},
{
"epoch": 0.10118577075098814,
"grad_norm": 8.3467435836792,
"learning_rate": 9.752821953287537e-06,
"loss": 0.2883,
"step": 160
},
{
"epoch": 0.1075098814229249,
"grad_norm": 6.815315246582031,
"learning_rate": 9.721054025049902e-06,
"loss": 0.3999,
"step": 170
},
{
"epoch": 0.11383399209486166,
"grad_norm": 6.288252353668213,
"learning_rate": 9.687424389154528e-06,
"loss": 0.3303,
"step": 180
},
{
"epoch": 0.12015810276679842,
"grad_norm": 5.876353740692139,
"learning_rate": 9.651946307163417e-06,
"loss": 0.3642,
"step": 190
},
{
"epoch": 0.12648221343873517,
"grad_norm": 7.05037784576416,
"learning_rate": 9.614633769557635e-06,
"loss": 0.2522,
"step": 200
},
{
"epoch": 0.13280632411067195,
"grad_norm": 4.7446980476379395,
"learning_rate": 9.57550149022027e-06,
"loss": 0.2671,
"step": 210
},
{
"epoch": 0.1391304347826087,
"grad_norm": 5.932438850402832,
"learning_rate": 9.534564900634146e-06,
"loss": 0.2612,
"step": 220
},
{
"epoch": 0.14545454545454545,
"grad_norm": 7.457898139953613,
"learning_rate": 9.491840143796554e-06,
"loss": 0.2939,
"step": 230
},
{
"epoch": 0.15177865612648223,
"grad_norm": 6.538613796234131,
"learning_rate": 9.44734406785338e-06,
"loss": 0.3139,
"step": 240
},
{
"epoch": 0.15810276679841898,
"grad_norm": 3.775620222091675,
"learning_rate": 9.401094219455202e-06,
"loss": 0.2799,
"step": 250
},
{
"epoch": 0.16442687747035573,
"grad_norm": 7.022812366485596,
"learning_rate": 9.353108836837907e-06,
"loss": 0.2448,
"step": 260
},
{
"epoch": 0.1707509881422925,
"grad_norm": 3.108099937438965,
"learning_rate": 9.30340684263059e-06,
"loss": 0.2596,
"step": 270
},
{
"epoch": 0.17707509881422925,
"grad_norm": 4.225621223449707,
"learning_rate": 9.252007836393593e-06,
"loss": 0.23,
"step": 280
},
{
"epoch": 0.183399209486166,
"grad_norm": 8.867532730102539,
"learning_rate": 9.198932086889567e-06,
"loss": 0.2782,
"step": 290
},
{
"epoch": 0.18972332015810275,
"grad_norm": 10.853947639465332,
"learning_rate": 9.14420052409067e-06,
"loss": 0.2611,
"step": 300
},
{
"epoch": 0.19604743083003953,
"grad_norm": 5.068134784698486,
"learning_rate": 9.087834730925002e-06,
"loss": 0.2399,
"step": 310
},
{
"epoch": 0.20237154150197628,
"grad_norm": 3.918778657913208,
"learning_rate": 9.029856934765565e-06,
"loss": 0.1992,
"step": 320
},
{
"epoch": 0.20869565217391303,
"grad_norm": 7.9009318351745605,
"learning_rate": 8.970289998665083e-06,
"loss": 0.287,
"step": 330
},
{
"epoch": 0.2150197628458498,
"grad_norm": 5.442112445831299,
"learning_rate": 8.90915741234015e-06,
"loss": 0.2279,
"step": 340
},
{
"epoch": 0.22134387351778656,
"grad_norm": 6.162675380706787,
"learning_rate": 8.846483282908267e-06,
"loss": 0.2894,
"step": 350
},
{
"epoch": 0.2276679841897233,
"grad_norm": 6.082408428192139,
"learning_rate": 8.782292325381397e-06,
"loss": 0.2405,
"step": 360
},
{
"epoch": 0.2339920948616601,
"grad_norm": 4.093995571136475,
"learning_rate": 8.716609852919808e-06,
"loss": 0.2769,
"step": 370
},
{
"epoch": 0.24031620553359684,
"grad_norm": 6.734269142150879,
"learning_rate": 8.649461766850053e-06,
"loss": 0.198,
"step": 380
},
{
"epoch": 0.2466403162055336,
"grad_norm": 7.3356523513793945,
"learning_rate": 8.580874546450992e-06,
"loss": 0.2377,
"step": 390
},
{
"epoch": 0.25296442687747034,
"grad_norm": 2.690619468688965,
"learning_rate": 8.510875238511911e-06,
"loss": 0.2073,
"step": 400
},
{
"epoch": 0.2592885375494071,
"grad_norm": 4.739744663238525,
"learning_rate": 8.439491446666858e-06,
"loss": 0.2374,
"step": 410
},
{
"epoch": 0.2656126482213439,
"grad_norm": 7.023416042327881,
"learning_rate": 8.366751320509383e-06,
"loss": 0.2858,
"step": 420
},
{
"epoch": 0.2719367588932806,
"grad_norm": 3.6838412284851074,
"learning_rate": 8.292683544491974e-06,
"loss": 0.1879,
"step": 430
},
{
"epoch": 0.2782608695652174,
"grad_norm": 8.874937057495117,
"learning_rate": 8.217317326614598e-06,
"loss": 0.1895,
"step": 440
},
{
"epoch": 0.2845849802371542,
"grad_norm": 4.901584148406982,
"learning_rate": 8.14068238690677e-06,
"loss": 0.2323,
"step": 450
},
{
"epoch": 0.2909090909090909,
"grad_norm": 7.248963356018066,
"learning_rate": 8.062808945707705e-06,
"loss": 0.2481,
"step": 460
},
{
"epoch": 0.2972332015810277,
"grad_norm": 5.835550308227539,
"learning_rate": 7.983727711749194e-06,
"loss": 0.2417,
"step": 470
},
{
"epoch": 0.30355731225296445,
"grad_norm": 4.541815280914307,
"learning_rate": 7.903469870045869e-06,
"loss": 0.2329,
"step": 480
},
{
"epoch": 0.3098814229249012,
"grad_norm": 2.072661876678467,
"learning_rate": 7.82206706959765e-06,
"loss": 0.208,
"step": 490
},
{
"epoch": 0.31620553359683795,
"grad_norm": 4.74345064163208,
"learning_rate": 7.739551410909254e-06,
"loss": 0.2855,
"step": 500
},
{
"epoch": 0.32252964426877473,
"grad_norm": 4.050395488739014,
"learning_rate": 7.6559554333316e-06,
"loss": 0.2206,
"step": 510
},
{
"epoch": 0.32885375494071145,
"grad_norm": 5.916479110717773,
"learning_rate": 7.571312102230218e-06,
"loss": 0.1384,
"step": 520
},
{
"epoch": 0.33517786561264823,
"grad_norm": 16.72198486328125,
"learning_rate": 7.485654795985626e-06,
"loss": 0.3058,
"step": 530
},
{
"epoch": 0.341501976284585,
"grad_norm": 4.175201892852783,
"learning_rate": 7.399017292830848e-06,
"loss": 0.1615,
"step": 540
},
{
"epoch": 0.34782608695652173,
"grad_norm": 7.27754020690918,
"learning_rate": 7.311433757531252e-06,
"loss": 0.2191,
"step": 550
},
{
"epoch": 0.3541501976284585,
"grad_norm": 5.6515302658081055,
"learning_rate": 7.222938727911974e-06,
"loss": 0.2436,
"step": 560
},
{
"epoch": 0.36047430830039523,
"grad_norm": 3.0964856147766113,
"learning_rate": 7.133567101238209e-06,
"loss": 0.1773,
"step": 570
},
{
"epoch": 0.366798418972332,
"grad_norm": 2.311328887939453,
"learning_rate": 7.043354120453771e-06,
"loss": 0.1859,
"step": 580
},
{
"epoch": 0.3731225296442688,
"grad_norm": 4.197350978851318,
"learning_rate": 6.9523353602833485e-06,
"loss": 0.1866,
"step": 590
},
{
"epoch": 0.3794466403162055,
"grad_norm": 8.702991485595703,
"learning_rate": 6.860546713203895e-06,
"loss": 0.2363,
"step": 600
},
{
"epoch": 0.3857707509881423,
"grad_norm": 4.5203094482421875,
"learning_rate": 6.768024375290747e-06,
"loss": 0.2037,
"step": 610
},
{
"epoch": 0.39209486166007906,
"grad_norm": 12.013908386230469,
"learning_rate": 6.6748048319440075e-06,
"loss": 0.2347,
"step": 620
},
{
"epoch": 0.3984189723320158,
"grad_norm": 5.178602695465088,
"learning_rate": 6.580924843500833e-06,
"loss": 0.2225,
"step": 630
},
{
"epoch": 0.40474308300395256,
"grad_norm": 2.2463157176971436,
"learning_rate": 6.486421430739309e-06,
"loss": 0.1776,
"step": 640
},
{
"epoch": 0.41106719367588934,
"grad_norm": 3.807896137237549,
"learning_rate": 6.391331860279623e-06,
"loss": 0.2285,
"step": 650
},
{
"epoch": 0.41739130434782606,
"grad_norm": 3.1088736057281494,
"learning_rate": 6.295693629888277e-06,
"loss": 0.2369,
"step": 660
},
{
"epoch": 0.42371541501976284,
"grad_norm": 2.071110725402832,
"learning_rate": 6.199544453691175e-06,
"loss": 0.2046,
"step": 670
},
{
"epoch": 0.4300395256916996,
"grad_norm": 2.81059193611145,
"learning_rate": 6.1029222473013705e-06,
"loss": 0.2245,
"step": 680
},
{
"epoch": 0.43636363636363634,
"grad_norm": 7.3108415603637695,
"learning_rate": 6.005865112867362e-06,
"loss": 0.2241,
"step": 690
},
{
"epoch": 0.4426877470355731,
"grad_norm": 4.426582336425781,
"learning_rate": 5.908411324047844e-06,
"loss": 0.1276,
"step": 700
},
{
"epoch": 0.4490118577075099,
"grad_norm": 4.832111358642578,
"learning_rate": 5.810599310918796e-06,
"loss": 0.2225,
"step": 710
},
{
"epoch": 0.4553359683794466,
"grad_norm": 5.967073917388916,
"learning_rate": 5.712467644818935e-06,
"loss": 0.2141,
"step": 720
},
{
"epoch": 0.4616600790513834,
"grad_norm": 5.369874000549316,
"learning_rate": 5.614055023139404e-06,
"loss": 0.2134,
"step": 730
},
{
"epoch": 0.4679841897233202,
"grad_norm": 2.798274278640747,
"learning_rate": 5.5154002540638096e-06,
"loss": 0.2272,
"step": 740
},
{
"epoch": 0.4743083003952569,
"grad_norm": 4.363977909088135,
"learning_rate": 5.416542241264524e-06,
"loss": 0.1628,
"step": 750
},
{
"epoch": 0.4806324110671937,
"grad_norm": 6.191987991333008,
"learning_rate": 5.317519968561367e-06,
"loss": 0.2072,
"step": 760
},
{
"epoch": 0.48695652173913045,
"grad_norm": 7.533758163452148,
"learning_rate": 5.2183724845486525e-06,
"loss": 0.2194,
"step": 770
},
{
"epoch": 0.4932806324110672,
"grad_norm": 7.126379013061523,
"learning_rate": 5.119138887196724e-06,
"loss": 0.1345,
"step": 780
},
{
"epoch": 0.49960474308300395,
"grad_norm": 2.0853822231292725,
"learning_rate": 5.019858308433979e-06,
"loss": 0.1466,
"step": 790
},
{
"epoch": 0.5059288537549407,
"grad_norm": 3.8167343139648438,
"learning_rate": 4.920569898715544e-06,
"loss": 0.1738,
"step": 800
},
{
"epoch": 0.5122529644268775,
"grad_norm": 3.872804880142212,
"learning_rate": 4.8213128115846136e-06,
"loss": 0.146,
"step": 810
},
{
"epoch": 0.5185770750988142,
"grad_norm": 3.171194553375244,
"learning_rate": 4.722126188232586e-06,
"loss": 0.1311,
"step": 820
},
{
"epoch": 0.524901185770751,
"grad_norm": 5.45040225982666,
"learning_rate": 4.623049142064066e-06,
"loss": 0.1682,
"step": 830
},
{
"epoch": 0.5312252964426878,
"grad_norm": 3.8132641315460205,
"learning_rate": 4.524120743272827e-06,
"loss": 0.2742,
"step": 840
},
{
"epoch": 0.5375494071146245,
"grad_norm": 2.825347661972046,
"learning_rate": 4.4253800034348e-06,
"loss": 0.1754,
"step": 850
},
{
"epoch": 0.5438735177865612,
"grad_norm": 3.77956485748291,
"learning_rate": 4.326865860124203e-06,
"loss": 0.2156,
"step": 860
},
{
"epoch": 0.5501976284584981,
"grad_norm": 6.331299781799316,
"learning_rate": 4.228617161558837e-06,
"loss": 0.1653,
"step": 870
},
{
"epoch": 0.5565217391304348,
"grad_norm": 5.832754135131836,
"learning_rate": 4.130672651280612e-06,
"loss": 0.1657,
"step": 880
},
{
"epoch": 0.5628458498023715,
"grad_norm": 4.60354471206665,
"learning_rate": 4.033070952877362e-06,
"loss": 0.1675,
"step": 890
},
{
"epoch": 0.5691699604743083,
"grad_norm": 1.3858933448791504,
"learning_rate": 3.935850554751964e-06,
"loss": 0.1499,
"step": 900
},
{
"epoch": 0.5754940711462451,
"grad_norm": 2.5932607650756836,
"learning_rate": 3.839049794944767e-06,
"loss": 0.1589,
"step": 910
},
{
"epoch": 0.5818181818181818,
"grad_norm": 3.6088171005249023,
"learning_rate": 3.7427068460152977e-06,
"loss": 0.1561,
"step": 920
},
{
"epoch": 0.5881422924901186,
"grad_norm": 5.195772647857666,
"learning_rate": 3.6468596999892415e-06,
"loss": 0.2044,
"step": 930
},
{
"epoch": 0.5944664031620553,
"grad_norm": 11.278207778930664,
"learning_rate": 3.551546153376611e-06,
"loss": 0.1862,
"step": 940
},
{
"epoch": 0.6007905138339921,
"grad_norm": 3.6617531776428223,
"learning_rate": 3.4568037922670135e-06,
"loss": 0.2161,
"step": 950
},
{
"epoch": 0.6071146245059289,
"grad_norm": 2.0994536876678467,
"learning_rate": 3.3626699775078884e-06,
"loss": 0.1423,
"step": 960
},
{
"epoch": 0.6134387351778656,
"grad_norm": 5.972179889678955,
"learning_rate": 3.269181829971581e-06,
"loss": 0.2189,
"step": 970
},
{
"epoch": 0.6197628458498023,
"grad_norm": 4.24259090423584,
"learning_rate": 3.1763762159170317e-06,
"loss": 0.1484,
"step": 980
},
{
"epoch": 0.6260869565217392,
"grad_norm": 6.183008193969727,
"learning_rate": 3.08428973245189e-06,
"loss": 0.2801,
"step": 990
},
{
"epoch": 0.6324110671936759,
"grad_norm": 3.033390522003174,
"learning_rate": 2.9929586931007494e-06,
"loss": 0.1291,
"step": 1000
},
{
"epoch": 0.6387351778656126,
"grad_norm": 3.0648672580718994,
"learning_rate": 2.9024191134852153e-06,
"loss": 0.1327,
"step": 1010
},
{
"epoch": 0.6450592885375495,
"grad_norm": 5.777024269104004,
"learning_rate": 2.8127066971214533e-06,
"loss": 0.1958,
"step": 1020
},
{
"epoch": 0.6513833992094862,
"grad_norm": 4.07697057723999,
"learning_rate": 2.723856821340806e-06,
"loss": 0.1842,
"step": 1030
},
{
"epoch": 0.6577075098814229,
"grad_norm": 3.3021464347839355,
"learning_rate": 2.6359045233390474e-06,
"loss": 0.1865,
"step": 1040
},
{
"epoch": 0.6640316205533597,
"grad_norm": 2.6843018531799316,
"learning_rate": 2.5488844863597613e-06,
"loss": 0.153,
"step": 1050
},
{
"epoch": 0.6703557312252965,
"grad_norm": 1.8710271120071411,
"learning_rate": 2.4628310260172838e-06,
"loss": 0.1591,
"step": 1060
},
{
"epoch": 0.6766798418972332,
"grad_norm": 1.834578037261963,
"learning_rate": 2.3777780767646556e-06,
"loss": 0.1354,
"step": 1070
},
{
"epoch": 0.68300395256917,
"grad_norm": 2.3767359256744385,
"learning_rate": 2.2937591785118407e-06,
"loss": 0.221,
"step": 1080
},
{
"epoch": 0.6893280632411067,
"grad_norm": 4.324508190155029,
"learning_rate": 2.210807463399554e-06,
"loss": 0.1483,
"step": 1090
},
{
"epoch": 0.6956521739130435,
"grad_norm": 2.8766467571258545,
"learning_rate": 2.128955642733877e-06,
"loss": 0.2214,
"step": 1100
},
{
"epoch": 0.7019762845849803,
"grad_norm": 1.6432983875274658,
"learning_rate": 2.0482359940868336e-06,
"loss": 0.1727,
"step": 1110
},
{
"epoch": 0.708300395256917,
"grad_norm": 2.7327544689178467,
"learning_rate": 1.9686803485680095e-06,
"loss": 0.1731,
"step": 1120
},
{
"epoch": 0.7146245059288537,
"grad_norm": 1.192838430404663,
"learning_rate": 1.890320078272212e-06,
"loss": 0.2139,
"step": 1130
},
{
"epoch": 0.7209486166007905,
"grad_norm": 3.182389259338379,
"learning_rate": 1.8131860839081634e-06,
"loss": 0.1351,
"step": 1140
},
{
"epoch": 0.7272727272727273,
"grad_norm": 2.460946798324585,
"learning_rate": 1.737308782613053e-06,
"loss": 0.1551,
"step": 1150
},
{
"epoch": 0.733596837944664,
"grad_norm": 17.547706604003906,
"learning_rate": 1.6627180959578143e-06,
"loss": 0.2043,
"step": 1160
},
{
"epoch": 0.7399209486166007,
"grad_norm": 2.621527671813965,
"learning_rate": 1.589443438147789e-06,
"loss": 0.1387,
"step": 1170
},
{
"epoch": 0.7462450592885376,
"grad_norm": 3.3238251209259033,
"learning_rate": 1.517513704423501e-06,
"loss": 0.1661,
"step": 1180
},
{
"epoch": 0.7525691699604743,
"grad_norm": 6.7597832679748535,
"learning_rate": 1.4469572596660486e-06,
"loss": 0.1849,
"step": 1190
},
{
"epoch": 0.758893280632411,
"grad_norm": 3.4403772354125977,
"learning_rate": 1.377801927211666e-06,
"loss": 0.1562,
"step": 1200
},
{
"epoch": 0.7652173913043478,
"grad_norm": 1.8223929405212402,
"learning_rate": 1.3100749778798166e-06,
"loss": 0.151,
"step": 1210
},
{
"epoch": 0.7715415019762846,
"grad_norm": 4.120046615600586,
"learning_rate": 1.2438031192191857e-06,
"loss": 0.1525,
"step": 1220
},
{
"epoch": 0.7778656126482213,
"grad_norm": 4.925788879394531,
"learning_rate": 1.1790124849757733e-06,
"loss": 0.1558,
"step": 1230
},
{
"epoch": 0.7841897233201581,
"grad_norm": 7.204642295837402,
"learning_rate": 1.1157286247872873e-06,
"loss": 0.1472,
"step": 1240
},
{
"epoch": 0.7905138339920948,
"grad_norm": 5.392803192138672,
"learning_rate": 1.0539764941078451e-06,
"loss": 0.1883,
"step": 1250
},
{
"epoch": 0.7968379446640316,
"grad_norm": 3.678210735321045,
"learning_rate": 9.937804443670136e-07,
"loss": 0.169,
"step": 1260
},
{
"epoch": 0.8031620553359684,
"grad_norm": 3.1506779193878174,
"learning_rate": 9.351642133670258e-07,
"loss": 0.1481,
"step": 1270
},
{
"epoch": 0.8094861660079051,
"grad_norm": 2.2924246788024902,
"learning_rate": 8.781509159219748e-07,
"loss": 0.1408,
"step": 1280
},
{
"epoch": 0.8158102766798419,
"grad_norm": 4.74252986907959,
"learning_rate": 8.227630347426868e-07,
"loss": 0.1778,
"step": 1290
},
{
"epoch": 0.8221343873517787,
"grad_norm": 3.5338618755340576,
"learning_rate": 7.690224115708573e-07,
"loss": 0.1264,
"step": 1300
},
{
"epoch": 0.8284584980237154,
"grad_norm": 3.4934120178222656,
"learning_rate": 7.16950238565941e-07,
"loss": 0.1194,
"step": 1310
},
{
"epoch": 0.8347826086956521,
"grad_norm": 1.5644766092300415,
"learning_rate": 6.66567049948203e-07,
"loss": 0.138,
"step": 1320
},
{
"epoch": 0.841106719367589,
"grad_norm": 5.495241165161133,
"learning_rate": 6.178927139012242e-07,
"loss": 0.1435,
"step": 1330
},
{
"epoch": 0.8474308300395257,
"grad_norm": 3.5011425018310547,
"learning_rate": 5.709464247370545e-07,
"loss": 0.1898,
"step": 1340
},
{
"epoch": 0.8537549407114624,
"grad_norm": 3.0744404792785645,
"learning_rate": 5.257466953270951e-07,
"loss": 0.1693,
"step": 1350
},
{
"epoch": 0.8600790513833992,
"grad_norm": 2.327395439147949,
"learning_rate": 4.823113498017024e-07,
"loss": 0.1486,
"step": 1360
},
{
"epoch": 0.866403162055336,
"grad_norm": 1.6566141843795776,
"learning_rate": 4.406575165213972e-07,
"loss": 0.118,
"step": 1370
},
{
"epoch": 0.8727272727272727,
"grad_norm": 9.42549991607666,
"learning_rate": 4.008016213224408e-07,
"loss": 0.2067,
"step": 1380
},
{
"epoch": 0.8790513833992095,
"grad_norm": 3.2364752292633057,
"learning_rate": 3.6275938103944477e-07,
"loss": 0.152,
"step": 1390
},
{
"epoch": 0.8853754940711462,
"grad_norm": 2.6394622325897217,
"learning_rate": 3.2654579730756695e-07,
"loss": 0.1441,
"step": 1400
},
{
"epoch": 0.891699604743083,
"grad_norm": 1.150625228881836,
"learning_rate": 2.921751506467479e-07,
"loss": 0.1464,
"step": 1410
},
{
"epoch": 0.8980237154150198,
"grad_norm": 3.9122185707092285,
"learning_rate": 2.596609948303097e-07,
"loss": 0.1242,
"step": 1420
},
{
"epoch": 0.9043478260869565,
"grad_norm": 4.106545925140381,
"learning_rate": 2.290161515401368e-07,
"loss": 0.1982,
"step": 1430
},
{
"epoch": 0.9106719367588932,
"grad_norm": 5.89559268951416,
"learning_rate": 2.0025270531055518e-07,
"loss": 0.2256,
"step": 1440
},
{
"epoch": 0.9169960474308301,
"grad_norm": 2.447909116744995,
"learning_rate": 1.7338199876289984e-07,
"loss": 0.1523,
"step": 1450
},
{
"epoch": 0.9233201581027668,
"grad_norm": 3.1286332607269287,
"learning_rate": 1.4841462813264763e-07,
"loss": 0.1204,
"step": 1460
},
{
"epoch": 0.9296442687747035,
"grad_norm": 5.63248348236084,
"learning_rate": 1.253604390908819e-07,
"loss": 0.1835,
"step": 1470
},
{
"epoch": 0.9359683794466404,
"grad_norm": 12.96162223815918,
"learning_rate": 1.0422852286173512e-07,
"loss": 0.1818,
"step": 1480
},
{
"epoch": 0.9422924901185771,
"grad_norm": 1.2593927383422852,
"learning_rate": 8.502721263734082e-08,
"loss": 0.1554,
"step": 1490
},
{
"epoch": 0.9486166007905138,
"grad_norm": 2.508932113647461,
"learning_rate": 6.776408029171178e-08,
"loss": 0.1142,
"step": 1500
},
{
"epoch": 0.9549407114624506,
"grad_norm": 4.1802215576171875,
"learning_rate": 5.244593339483406e-08,
"loss": 0.1109,
"step": 1510
},
{
"epoch": 0.9612648221343874,
"grad_norm": 9.903648376464844,
"learning_rate": 3.907881252816048e-08,
"loss": 0.1663,
"step": 1520
},
{
"epoch": 0.9675889328063241,
"grad_norm": 5.05954647064209,
"learning_rate": 2.7667988902557174e-08,
"loss": 0.1694,
"step": 1530
},
{
"epoch": 0.9739130434782609,
"grad_norm": 3.208235025405884,
"learning_rate": 1.821796227964856e-08,
"loss": 0.1412,
"step": 1540
},
{
"epoch": 0.9802371541501976,
"grad_norm": 5.830465316772461,
"learning_rate": 1.0732459197371758e-08,
"loss": 0.1584,
"step": 1550
},
{
"epoch": 0.9865612648221344,
"grad_norm": 7.674189567565918,
"learning_rate": 5.214431500449868e-09,
"loss": 0.1613,
"step": 1560
},
{
"epoch": 0.9928853754940712,
"grad_norm": 5.355103492736816,
"learning_rate": 1.6660551763564292e-09,
"loss": 0.149,
"step": 1570
},
{
"epoch": 0.9992094861660079,
"grad_norm": 2.4179863929748535,
"learning_rate": 8.87294972318209e-11,
"loss": 0.174,
"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
}