M3D-RAD-Gemma3-4B-A1 / trainer_state.json
fadefade's picture
Add files using upload-large-folder tool
80c1007 verified
Raw
History Blame Contribute Delete
141 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 370,
"global_step": 3695,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006765899864682003,
"grad_norm": 6.708961009979248,
"learning_rate": 1.801801801801802e-06,
"loss": 0.3380268096923828,
"step": 5
},
{
"epoch": 0.013531799729364006,
"grad_norm": 6.754238128662109,
"learning_rate": 4.0540540540540545e-06,
"loss": 0.30354058742523193,
"step": 10
},
{
"epoch": 0.02029769959404601,
"grad_norm": 5.660676002502441,
"learning_rate": 6.306306306306306e-06,
"loss": 0.32503542900085447,
"step": 15
},
{
"epoch": 0.02706359945872801,
"grad_norm": 4.619103908538818,
"learning_rate": 8.558558558558558e-06,
"loss": 0.2942976951599121,
"step": 20
},
{
"epoch": 0.03382949932341001,
"grad_norm": 7.728912353515625,
"learning_rate": 1.0810810810810812e-05,
"loss": 0.2309746265411377,
"step": 25
},
{
"epoch": 0.04059539918809202,
"grad_norm": 6.946638107299805,
"learning_rate": 1.3063063063063064e-05,
"loss": 0.3821186304092407,
"step": 30
},
{
"epoch": 0.04736129905277402,
"grad_norm": 7.483209609985352,
"learning_rate": 1.5315315315315316e-05,
"loss": 0.272556734085083,
"step": 35
},
{
"epoch": 0.05412719891745602,
"grad_norm": 5.25261116027832,
"learning_rate": 1.756756756756757e-05,
"loss": 0.3210551023483276,
"step": 40
},
{
"epoch": 0.06089309878213803,
"grad_norm": 5.864253997802734,
"learning_rate": 1.981981981981982e-05,
"loss": 0.36204843521118163,
"step": 45
},
{
"epoch": 0.06765899864682003,
"grad_norm": 5.3668670654296875,
"learning_rate": 2.2072072072072073e-05,
"loss": 0.2810213565826416,
"step": 50
},
{
"epoch": 0.07442489851150202,
"grad_norm": 6.58739709854126,
"learning_rate": 2.4324324324324327e-05,
"loss": 0.35600206851959226,
"step": 55
},
{
"epoch": 0.08119079837618404,
"grad_norm": 4.801163196563721,
"learning_rate": 2.6576576576576577e-05,
"loss": 0.31467452049255373,
"step": 60
},
{
"epoch": 0.08795669824086604,
"grad_norm": 5.191904544830322,
"learning_rate": 2.882882882882883e-05,
"loss": 0.28929872512817384,
"step": 65
},
{
"epoch": 0.09472259810554803,
"grad_norm": 5.917762279510498,
"learning_rate": 3.108108108108108e-05,
"loss": 0.3355743408203125,
"step": 70
},
{
"epoch": 0.10148849797023005,
"grad_norm": 4.782976150512695,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.27232046127319337,
"step": 75
},
{
"epoch": 0.10825439783491204,
"grad_norm": 11.905159950256348,
"learning_rate": 3.558558558558558e-05,
"loss": 0.3356289863586426,
"step": 80
},
{
"epoch": 0.11502029769959404,
"grad_norm": 4.453586578369141,
"learning_rate": 3.783783783783784e-05,
"loss": 0.2907212734222412,
"step": 85
},
{
"epoch": 0.12178619756427606,
"grad_norm": 5.226840019226074,
"learning_rate": 4.0090090090090096e-05,
"loss": 0.3151070594787598,
"step": 90
},
{
"epoch": 0.12855209742895804,
"grad_norm": 4.185660362243652,
"learning_rate": 4.234234234234234e-05,
"loss": 0.23938796520233155,
"step": 95
},
{
"epoch": 0.13531799729364005,
"grad_norm": 3.919585704803467,
"learning_rate": 4.4594594594594596e-05,
"loss": 0.22664246559143067,
"step": 100
},
{
"epoch": 0.14208389715832206,
"grad_norm": 3.596623659133911,
"learning_rate": 4.684684684684685e-05,
"loss": 0.3125948905944824,
"step": 105
},
{
"epoch": 0.14884979702300405,
"grad_norm": 7.164392948150635,
"learning_rate": 4.90990990990991e-05,
"loss": 0.40825443267822265,
"step": 110
},
{
"epoch": 0.15561569688768606,
"grad_norm": 6.525625705718994,
"learning_rate": 4.999991355974974e-05,
"loss": 0.3195007801055908,
"step": 115
},
{
"epoch": 0.16238159675236807,
"grad_norm": 2.3961687088012695,
"learning_rate": 4.999938531594062e-05,
"loss": 0.2873862266540527,
"step": 120
},
{
"epoch": 0.16914749661705006,
"grad_norm": 7.02092170715332,
"learning_rate": 4.999837686081832e-05,
"loss": 0.29604496955871584,
"step": 125
},
{
"epoch": 0.17591339648173207,
"grad_norm": 4.478824138641357,
"learning_rate": 4.9996888213754195e-05,
"loss": 0.29402358531951905,
"step": 130
},
{
"epoch": 0.18267929634641408,
"grad_norm": 3.910637617111206,
"learning_rate": 4.999491940334355e-05,
"loss": 0.2536696195602417,
"step": 135
},
{
"epoch": 0.18944519621109607,
"grad_norm": 4.6856465339660645,
"learning_rate": 4.9992470467405104e-05,
"loss": 0.2898806095123291,
"step": 140
},
{
"epoch": 0.19621109607577808,
"grad_norm": 3.10506534576416,
"learning_rate": 4.99895414529803e-05,
"loss": 0.3189587116241455,
"step": 145
},
{
"epoch": 0.2029769959404601,
"grad_norm": 2.7629287242889404,
"learning_rate": 4.9986132416332353e-05,
"loss": 0.2612765789031982,
"step": 150
},
{
"epoch": 0.20974289580514208,
"grad_norm": 4.227510929107666,
"learning_rate": 4.998224342294518e-05,
"loss": 0.3242856502532959,
"step": 155
},
{
"epoch": 0.2165087956698241,
"grad_norm": 3.744783401489258,
"learning_rate": 4.9977874547522175e-05,
"loss": 0.21590490341186525,
"step": 160
},
{
"epoch": 0.2232746955345061,
"grad_norm": 2.8258867263793945,
"learning_rate": 4.997302587398473e-05,
"loss": 0.2537590980529785,
"step": 165
},
{
"epoch": 0.23004059539918809,
"grad_norm": 4.918529033660889,
"learning_rate": 4.996769749547065e-05,
"loss": 0.3003899574279785,
"step": 170
},
{
"epoch": 0.2368064952638701,
"grad_norm": 2.7496345043182373,
"learning_rate": 4.9961889514332375e-05,
"loss": 0.30573205947875975,
"step": 175
},
{
"epoch": 0.2435723951285521,
"grad_norm": 2.909468412399292,
"learning_rate": 4.9955602042134955e-05,
"loss": 0.32769505977630614,
"step": 180
},
{
"epoch": 0.2503382949932341,
"grad_norm": 1.5789233446121216,
"learning_rate": 4.9948835199653974e-05,
"loss": 0.2218461275100708,
"step": 185
},
{
"epoch": 0.2571041948579161,
"grad_norm": 3.51835298538208,
"learning_rate": 4.994158911687321e-05,
"loss": 0.27026145458221434,
"step": 190
},
{
"epoch": 0.2638700947225981,
"grad_norm": 4.7750163078308105,
"learning_rate": 4.993386393298212e-05,
"loss": 0.2207875967025757,
"step": 195
},
{
"epoch": 0.2706359945872801,
"grad_norm": 2.666891098022461,
"learning_rate": 4.9925659796373184e-05,
"loss": 0.21574645042419432,
"step": 200
},
{
"epoch": 0.2774018944519621,
"grad_norm": 3.404348134994507,
"learning_rate": 4.991697686463903e-05,
"loss": 0.26267313957214355,
"step": 205
},
{
"epoch": 0.28416779431664413,
"grad_norm": 2.4391016960144043,
"learning_rate": 4.990781530456945e-05,
"loss": 0.2689904928207397,
"step": 210
},
{
"epoch": 0.29093369418132614,
"grad_norm": 4.201110363006592,
"learning_rate": 4.9898175292148177e-05,
"loss": 0.26420440673828127,
"step": 215
},
{
"epoch": 0.2976995940460081,
"grad_norm": 4.51666784286499,
"learning_rate": 4.988805701254946e-05,
"loss": 0.28138198852539065,
"step": 220
},
{
"epoch": 0.3044654939106901,
"grad_norm": 2.9034829139709473,
"learning_rate": 4.987746066013459e-05,
"loss": 0.2245555877685547,
"step": 225
},
{
"epoch": 0.3112313937753721,
"grad_norm": 3.982085943222046,
"learning_rate": 4.986638643844809e-05,
"loss": 0.27320256233215334,
"step": 230
},
{
"epoch": 0.31799729364005414,
"grad_norm": 2.4032037258148193,
"learning_rate": 4.9854834560213865e-05,
"loss": 0.24287850856781007,
"step": 235
},
{
"epoch": 0.32476319350473615,
"grad_norm": 2.052222490310669,
"learning_rate": 4.984280524733107e-05,
"loss": 0.2938188076019287,
"step": 240
},
{
"epoch": 0.33152909336941816,
"grad_norm": 2.6662299633026123,
"learning_rate": 4.9830298730869864e-05,
"loss": 0.2657630920410156,
"step": 245
},
{
"epoch": 0.3382949932341001,
"grad_norm": 3.018406391143799,
"learning_rate": 4.9817315251067e-05,
"loss": 0.24283986091613768,
"step": 250
},
{
"epoch": 0.34506089309878213,
"grad_norm": 4.130184173583984,
"learning_rate": 4.980385505732115e-05,
"loss": 0.3305687427520752,
"step": 255
},
{
"epoch": 0.35182679296346414,
"grad_norm": 2.634552240371704,
"learning_rate": 4.978991840818816e-05,
"loss": 0.271871542930603,
"step": 260
},
{
"epoch": 0.35859269282814615,
"grad_norm": 2.4888367652893066,
"learning_rate": 4.977550557137609e-05,
"loss": 0.257403564453125,
"step": 265
},
{
"epoch": 0.36535859269282817,
"grad_norm": 2.5016252994537354,
"learning_rate": 4.9760616823740025e-05,
"loss": 0.25123617649078367,
"step": 270
},
{
"epoch": 0.3721244925575101,
"grad_norm": 2.4973812103271484,
"learning_rate": 4.9745252451276815e-05,
"loss": 0.24676234722137452,
"step": 275
},
{
"epoch": 0.37889039242219213,
"grad_norm": 1.979268193244934,
"learning_rate": 4.972941274911953e-05,
"loss": 0.18489110469818115,
"step": 280
},
{
"epoch": 0.38565629228687415,
"grad_norm": 3.698150157928467,
"learning_rate": 4.9713098021531806e-05,
"loss": 0.2770824432373047,
"step": 285
},
{
"epoch": 0.39242219215155616,
"grad_norm": 1.4433162212371826,
"learning_rate": 4.969630858190203e-05,
"loss": 0.22050118446350098,
"step": 290
},
{
"epoch": 0.39918809201623817,
"grad_norm": 3.9170804023742676,
"learning_rate": 4.9679044752737286e-05,
"loss": 0.2317894220352173,
"step": 295
},
{
"epoch": 0.4059539918809202,
"grad_norm": 2.1015920639038086,
"learning_rate": 4.9661306865657164e-05,
"loss": 0.20321197509765626,
"step": 300
},
{
"epoch": 0.41271989174560214,
"grad_norm": 2.286485433578491,
"learning_rate": 4.9643095261387414e-05,
"loss": 0.3368730306625366,
"step": 305
},
{
"epoch": 0.41948579161028415,
"grad_norm": 3.262737512588501,
"learning_rate": 4.962441028975336e-05,
"loss": 0.39606666564941406,
"step": 310
},
{
"epoch": 0.42625169147496617,
"grad_norm": 1.6825315952301025,
"learning_rate": 4.9605252309673226e-05,
"loss": 0.19189832210540772,
"step": 315
},
{
"epoch": 0.4330175913396482,
"grad_norm": 3.355905771255493,
"learning_rate": 4.9585621689151216e-05,
"loss": 0.26699066162109375,
"step": 320
},
{
"epoch": 0.4397834912043302,
"grad_norm": 1.9652124643325806,
"learning_rate": 4.956551880527043e-05,
"loss": 0.2927189826965332,
"step": 325
},
{
"epoch": 0.4465493910690122,
"grad_norm": 1.9714268445968628,
"learning_rate": 4.954494404418566e-05,
"loss": 0.20982046127319337,
"step": 330
},
{
"epoch": 0.45331529093369416,
"grad_norm": 4.401551723480225,
"learning_rate": 4.952389780111595e-05,
"loss": 0.31375908851623535,
"step": 335
},
{
"epoch": 0.46008119079837617,
"grad_norm": 2.7989892959594727,
"learning_rate": 4.950238048033699e-05,
"loss": 0.2211078643798828,
"step": 340
},
{
"epoch": 0.4668470906630582,
"grad_norm": 1.8418807983398438,
"learning_rate": 4.948039249517339e-05,
"loss": 0.27459444999694826,
"step": 345
},
{
"epoch": 0.4736129905277402,
"grad_norm": 2.5638668537139893,
"learning_rate": 4.9457934267990693e-05,
"loss": 0.2483287811279297,
"step": 350
},
{
"epoch": 0.4803788903924222,
"grad_norm": 2.9532172679901123,
"learning_rate": 4.94350062301873e-05,
"loss": 0.25247581005096437,
"step": 355
},
{
"epoch": 0.4871447902571042,
"grad_norm": 1.9480071067810059,
"learning_rate": 4.941160882218617e-05,
"loss": 0.34073171615600584,
"step": 360
},
{
"epoch": 0.4939106901217862,
"grad_norm": 3.35947585105896,
"learning_rate": 4.9387742493426355e-05,
"loss": 0.1844736099243164,
"step": 365
},
{
"epoch": 0.5006765899864682,
"grad_norm": 3.0802876949310303,
"learning_rate": 4.936340770235436e-05,
"loss": 0.2683614730834961,
"step": 370
},
{
"epoch": 0.5006765899864682,
"eval_accuracy": 0.8558597554377232,
"eval_loss": 0.3399568796157837,
"eval_runtime": 28.9163,
"eval_samples_per_second": 70.825,
"eval_steps_per_second": 4.427,
"step": 370
},
{
"epoch": 0.5074424898511503,
"grad_norm": 3.6506221294403076,
"learning_rate": 4.933860491641537e-05,
"loss": 0.24743883609771727,
"step": 375
},
{
"epoch": 0.5142083897158322,
"grad_norm": 2.7094950675964355,
"learning_rate": 4.9313334612044216e-05,
"loss": 0.20601317882537842,
"step": 380
},
{
"epoch": 0.5209742895805142,
"grad_norm": 2.782653331756592,
"learning_rate": 4.928759727465629e-05,
"loss": 0.3552024602890015,
"step": 385
},
{
"epoch": 0.5277401894451962,
"grad_norm": 2.0927574634552,
"learning_rate": 4.926139339863818e-05,
"loss": 0.16139503717422485,
"step": 390
},
{
"epoch": 0.5345060893098782,
"grad_norm": 2.3310742378234863,
"learning_rate": 4.923472348733815e-05,
"loss": 0.375275993347168,
"step": 395
},
{
"epoch": 0.5412719891745602,
"grad_norm": 2.2028820514678955,
"learning_rate": 4.9207588053056545e-05,
"loss": 0.2704309463500977,
"step": 400
},
{
"epoch": 0.5480378890392422,
"grad_norm": 2.372511625289917,
"learning_rate": 4.917998761703585e-05,
"loss": 0.30387301445007325,
"step": 405
},
{
"epoch": 0.5548037889039242,
"grad_norm": 1.546380639076233,
"learning_rate": 4.9151922709450795e-05,
"loss": 0.23326597213745118,
"step": 410
},
{
"epoch": 0.5615696887686062,
"grad_norm": 1.753524899482727,
"learning_rate": 4.912339386939806e-05,
"loss": 0.22853460311889648,
"step": 415
},
{
"epoch": 0.5683355886332883,
"grad_norm": 1.8898591995239258,
"learning_rate": 4.9094401644886e-05,
"loss": 0.18049622774124147,
"step": 420
},
{
"epoch": 0.5751014884979703,
"grad_norm": 2.374713182449341,
"learning_rate": 4.906494659282407e-05,
"loss": 0.2383967399597168,
"step": 425
},
{
"epoch": 0.5818673883626523,
"grad_norm": 1.690330147743225,
"learning_rate": 4.903502927901216e-05,
"loss": 0.18138002157211303,
"step": 430
},
{
"epoch": 0.5886332882273342,
"grad_norm": 2.2419376373291016,
"learning_rate": 4.900465027812969e-05,
"loss": 0.21561579704284667,
"step": 435
},
{
"epoch": 0.5953991880920162,
"grad_norm": 2.2026755809783936,
"learning_rate": 4.8973810173724624e-05,
"loss": 0.23320369720458983,
"step": 440
},
{
"epoch": 0.6021650879566982,
"grad_norm": 2.5355589389801025,
"learning_rate": 4.894250955820218e-05,
"loss": 0.2214221954345703,
"step": 445
},
{
"epoch": 0.6089309878213802,
"grad_norm": 2.4363200664520264,
"learning_rate": 4.8910749032813555e-05,
"loss": 0.24546003341674805,
"step": 450
},
{
"epoch": 0.6156968876860622,
"grad_norm": 2.513256311416626,
"learning_rate": 4.8878529207644275e-05,
"loss": 0.20654470920562745,
"step": 455
},
{
"epoch": 0.6224627875507442,
"grad_norm": 3.2197139263153076,
"learning_rate": 4.884585070160255e-05,
"loss": 0.28106026649475097,
"step": 460
},
{
"epoch": 0.6292286874154263,
"grad_norm": 1.973872423171997,
"learning_rate": 4.881271414240733e-05,
"loss": 0.24674997329711915,
"step": 465
},
{
"epoch": 0.6359945872801083,
"grad_norm": 1.55397367477417,
"learning_rate": 4.877912016657632e-05,
"loss": 0.254152250289917,
"step": 470
},
{
"epoch": 0.6427604871447903,
"grad_norm": 1.9519964456558228,
"learning_rate": 4.8745069419413646e-05,
"loss": 0.14674036502838134,
"step": 475
},
{
"epoch": 0.6495263870094723,
"grad_norm": 1.71841299533844,
"learning_rate": 4.871056255499757e-05,
"loss": 0.24175169467926025,
"step": 480
},
{
"epoch": 0.6562922868741543,
"grad_norm": 2.4556474685668945,
"learning_rate": 4.867560023616785e-05,
"loss": 0.2484445571899414,
"step": 485
},
{
"epoch": 0.6630581867388363,
"grad_norm": 1.6898027658462524,
"learning_rate": 4.8640183134513036e-05,
"loss": 0.197585391998291,
"step": 490
},
{
"epoch": 0.6698240866035182,
"grad_norm": 2.166822671890259,
"learning_rate": 4.860431193035758e-05,
"loss": 0.24004759788513183,
"step": 495
},
{
"epoch": 0.6765899864682002,
"grad_norm": 1.966162919998169,
"learning_rate": 4.856798731274873e-05,
"loss": 0.27218830585479736,
"step": 500
},
{
"epoch": 0.6833558863328822,
"grad_norm": 2.945369243621826,
"learning_rate": 4.853120997944335e-05,
"loss": 0.27289423942565916,
"step": 505
},
{
"epoch": 0.6901217861975643,
"grad_norm": 3.2015297412872314,
"learning_rate": 4.8493980636894473e-05,
"loss": 0.2509544849395752,
"step": 510
},
{
"epoch": 0.6968876860622463,
"grad_norm": 3.208461046218872,
"learning_rate": 4.8456300000237724e-05,
"loss": 0.2308948278427124,
"step": 515
},
{
"epoch": 0.7036535859269283,
"grad_norm": 1.398843765258789,
"learning_rate": 4.8418168793277644e-05,
"loss": 0.14355896711349486,
"step": 520
},
{
"epoch": 0.7104194857916103,
"grad_norm": 1.889447808265686,
"learning_rate": 4.8379587748473696e-05,
"loss": 0.2273927927017212,
"step": 525
},
{
"epoch": 0.7171853856562923,
"grad_norm": 1.8295615911483765,
"learning_rate": 4.834055760692628e-05,
"loss": 0.2629696846008301,
"step": 530
},
{
"epoch": 0.7239512855209743,
"grad_norm": 4.660758018493652,
"learning_rate": 4.830107911836245e-05,
"loss": 0.25275726318359376,
"step": 535
},
{
"epoch": 0.7307171853856563,
"grad_norm": 3.817150354385376,
"learning_rate": 4.8261153041121535e-05,
"loss": 0.24568512439727783,
"step": 540
},
{
"epoch": 0.7374830852503383,
"grad_norm": 1.4462109804153442,
"learning_rate": 4.8220780142140534e-05,
"loss": 0.21610636711120607,
"step": 545
},
{
"epoch": 0.7442489851150202,
"grad_norm": 2.598623275756836,
"learning_rate": 4.817996119693944e-05,
"loss": 0.29136559963226316,
"step": 550
},
{
"epoch": 0.7510148849797023,
"grad_norm": 1.553445816040039,
"learning_rate": 4.81386969896063e-05,
"loss": 0.222825288772583,
"step": 555
},
{
"epoch": 0.7577807848443843,
"grad_norm": 2.352839946746826,
"learning_rate": 4.8096988312782174e-05,
"loss": 0.29239563941955565,
"step": 560
},
{
"epoch": 0.7645466847090663,
"grad_norm": 2.579105854034424,
"learning_rate": 4.805483596764589e-05,
"loss": 0.2721230506896973,
"step": 565
},
{
"epoch": 0.7713125845737483,
"grad_norm": 2.563720941543579,
"learning_rate": 4.8012240763898686e-05,
"loss": 0.3364142894744873,
"step": 570
},
{
"epoch": 0.7780784844384303,
"grad_norm": 3.1388018131256104,
"learning_rate": 4.7969203519748615e-05,
"loss": 0.23970434665679932,
"step": 575
},
{
"epoch": 0.7848443843031123,
"grad_norm": 2.5504119396209717,
"learning_rate": 4.792572506189489e-05,
"loss": 0.2797069549560547,
"step": 580
},
{
"epoch": 0.7916102841677943,
"grad_norm": 1.2897508144378662,
"learning_rate": 4.788180622551191e-05,
"loss": 0.21211330890655516,
"step": 585
},
{
"epoch": 0.7983761840324763,
"grad_norm": 1.872567057609558,
"learning_rate": 4.783744785423333e-05,
"loss": 0.21291179656982423,
"step": 590
},
{
"epoch": 0.8051420838971584,
"grad_norm": 1.4606621265411377,
"learning_rate": 4.779265080013575e-05,
"loss": 0.24714994430541992,
"step": 595
},
{
"epoch": 0.8119079837618404,
"grad_norm": 4.170928001403809,
"learning_rate": 4.774741592372241e-05,
"loss": 0.25944693088531495,
"step": 600
},
{
"epoch": 0.8186738836265224,
"grad_norm": 2.2985825538635254,
"learning_rate": 4.770174409390668e-05,
"loss": 0.2784742832183838,
"step": 605
},
{
"epoch": 0.8254397834912043,
"grad_norm": 1.605530023574829,
"learning_rate": 4.7655636187995266e-05,
"loss": 0.2133476972579956,
"step": 610
},
{
"epoch": 0.8322056833558863,
"grad_norm": 2.021308660507202,
"learning_rate": 4.760909309167149e-05,
"loss": 0.2779200077056885,
"step": 615
},
{
"epoch": 0.8389715832205683,
"grad_norm": 3.999857187271118,
"learning_rate": 4.7562115698978186e-05,
"loss": 0.2434375762939453,
"step": 620
},
{
"epoch": 0.8457374830852503,
"grad_norm": 1.4456400871276855,
"learning_rate": 4.751470491230056e-05,
"loss": 0.2559937000274658,
"step": 625
},
{
"epoch": 0.8525033829499323,
"grad_norm": 2.1214683055877686,
"learning_rate": 4.746686164234885e-05,
"loss": 0.2085390567779541,
"step": 630
},
{
"epoch": 0.8592692828146143,
"grad_norm": 2.249709129333496,
"learning_rate": 4.741858680814083e-05,
"loss": 0.21808948516845703,
"step": 635
},
{
"epoch": 0.8660351826792964,
"grad_norm": 1.7097965478897095,
"learning_rate": 4.7369881336984153e-05,
"loss": 0.24430737495422364,
"step": 640
},
{
"epoch": 0.8728010825439784,
"grad_norm": 2.684420108795166,
"learning_rate": 4.7320746164458563e-05,
"loss": 0.21013333797454833,
"step": 645
},
{
"epoch": 0.8795669824086604,
"grad_norm": 3.2410199642181396,
"learning_rate": 4.7271182234397896e-05,
"loss": 0.15250587463378906,
"step": 650
},
{
"epoch": 0.8863328822733424,
"grad_norm": 2.4872889518737793,
"learning_rate": 4.722119049887196e-05,
"loss": 0.296546483039856,
"step": 655
},
{
"epoch": 0.8930987821380244,
"grad_norm": 3.6662309169769287,
"learning_rate": 4.717077191816823e-05,
"loss": 0.2242109775543213,
"step": 660
},
{
"epoch": 0.8998646820027063,
"grad_norm": 2.0327024459838867,
"learning_rate": 4.711992746077345e-05,
"loss": 0.2253121852874756,
"step": 665
},
{
"epoch": 0.9066305818673883,
"grad_norm": 2.6694586277008057,
"learning_rate": 4.706865810335497e-05,
"loss": 0.21453540325164794,
"step": 670
},
{
"epoch": 0.9133964817320703,
"grad_norm": 3.337958574295044,
"learning_rate": 4.701696483074202e-05,
"loss": 0.2270751953125,
"step": 675
},
{
"epoch": 0.9201623815967523,
"grad_norm": 2.4391632080078125,
"learning_rate": 4.696484863590679e-05,
"loss": 0.24238224029541017,
"step": 680
},
{
"epoch": 0.9269282814614344,
"grad_norm": 2.061091423034668,
"learning_rate": 4.6912310519945356e-05,
"loss": 0.2503541946411133,
"step": 685
},
{
"epoch": 0.9336941813261164,
"grad_norm": 1.3907153606414795,
"learning_rate": 4.6859351492058434e-05,
"loss": 0.29397246837615965,
"step": 690
},
{
"epoch": 0.9404600811907984,
"grad_norm": 2.2361032962799072,
"learning_rate": 4.6805972569532007e-05,
"loss": 0.2425755023956299,
"step": 695
},
{
"epoch": 0.9472259810554804,
"grad_norm": 2.3671493530273438,
"learning_rate": 4.6752174777717786e-05,
"loss": 0.2526777505874634,
"step": 700
},
{
"epoch": 0.9539918809201624,
"grad_norm": 1.2576649188995361,
"learning_rate": 4.669795915001351e-05,
"loss": 0.24289484024047853,
"step": 705
},
{
"epoch": 0.9607577807848444,
"grad_norm": 1.857110619544983,
"learning_rate": 4.66433267278431e-05,
"loss": 0.1650902271270752,
"step": 710
},
{
"epoch": 0.9675236806495264,
"grad_norm": 2.46406888961792,
"learning_rate": 4.6588278560636636e-05,
"loss": 0.2542042970657349,
"step": 715
},
{
"epoch": 0.9742895805142084,
"grad_norm": 1.6436994075775146,
"learning_rate": 4.653281570581023e-05,
"loss": 0.27213177680969236,
"step": 720
},
{
"epoch": 0.9810554803788903,
"grad_norm": 1.781708002090454,
"learning_rate": 4.647693922874569e-05,
"loss": 0.193695330619812,
"step": 725
},
{
"epoch": 0.9878213802435724,
"grad_norm": 1.9310158491134644,
"learning_rate": 4.642065020277006e-05,
"loss": 0.21228816509246826,
"step": 730
},
{
"epoch": 0.9945872801082544,
"grad_norm": 1.6782934665679932,
"learning_rate": 4.6363949709135014e-05,
"loss": 0.18696770668029786,
"step": 735
},
{
"epoch": 1.0013531799729365,
"grad_norm": 1.2074904441833496,
"learning_rate": 4.630683883699607e-05,
"loss": 0.23858895301818847,
"step": 740
},
{
"epoch": 1.0013531799729365,
"eval_accuracy": 0.8647332539768423,
"eval_loss": 0.3140738904476166,
"eval_runtime": 28.9416,
"eval_samples_per_second": 70.763,
"eval_steps_per_second": 4.423,
"step": 740
},
{
"epoch": 1.0081190798376185,
"grad_norm": 2.295377254486084,
"learning_rate": 4.6249318683391685e-05,
"loss": 0.2022010326385498,
"step": 745
},
{
"epoch": 1.0148849797023005,
"grad_norm": 1.6932891607284546,
"learning_rate": 4.619139035322215e-05,
"loss": 0.18071336746215821,
"step": 750
},
{
"epoch": 1.0216508795669823,
"grad_norm": 1.825318694114685,
"learning_rate": 4.613305495922841e-05,
"loss": 0.1911982774734497,
"step": 755
},
{
"epoch": 1.0284167794316643,
"grad_norm": 2.134507417678833,
"learning_rate": 4.6074313621970665e-05,
"loss": 0.23797287940979003,
"step": 760
},
{
"epoch": 1.0351826792963463,
"grad_norm": 2.610055446624756,
"learning_rate": 4.6015167469806854e-05,
"loss": 0.14118094444274903,
"step": 765
},
{
"epoch": 1.0419485791610283,
"grad_norm": 2.076680898666382,
"learning_rate": 4.595561763887095e-05,
"loss": 0.22451024055480956,
"step": 770
},
{
"epoch": 1.0487144790257104,
"grad_norm": 1.1106986999511719,
"learning_rate": 4.58956652730512e-05,
"loss": 0.19099733829498292,
"step": 775
},
{
"epoch": 1.0554803788903924,
"grad_norm": 1.8168954849243164,
"learning_rate": 4.583531152396807e-05,
"loss": 0.24614920616149902,
"step": 780
},
{
"epoch": 1.0622462787550744,
"grad_norm": 2.0116636753082275,
"learning_rate": 4.5774557550952204e-05,
"loss": 0.1600014567375183,
"step": 785
},
{
"epoch": 1.0690121786197564,
"grad_norm": 1.84378981590271,
"learning_rate": 4.5713404521022095e-05,
"loss": 0.2103560209274292,
"step": 790
},
{
"epoch": 1.0757780784844384,
"grad_norm": 1.42452073097229,
"learning_rate": 4.56518536088617e-05,
"loss": 0.19004445075988768,
"step": 795
},
{
"epoch": 1.0825439783491204,
"grad_norm": 1.9308594465255737,
"learning_rate": 4.558990599679787e-05,
"loss": 0.17624146938323976,
"step": 800
},
{
"epoch": 1.0893098782138024,
"grad_norm": 2.370720863342285,
"learning_rate": 4.552756287477762e-05,
"loss": 0.1937591552734375,
"step": 805
},
{
"epoch": 1.0960757780784844,
"grad_norm": 2.1250479221343994,
"learning_rate": 4.546482544034529e-05,
"loss": 0.19167513847351075,
"step": 810
},
{
"epoch": 1.1028416779431665,
"grad_norm": 2.417142868041992,
"learning_rate": 4.540169489861956e-05,
"loss": 0.199344003200531,
"step": 815
},
{
"epoch": 1.1096075778078485,
"grad_norm": 1.273617148399353,
"learning_rate": 4.533817246227024e-05,
"loss": 0.18206419944763183,
"step": 820
},
{
"epoch": 1.1163734776725305,
"grad_norm": 1.907364010810852,
"learning_rate": 4.527425935149504e-05,
"loss": 0.2041105270385742,
"step": 825
},
{
"epoch": 1.1231393775372125,
"grad_norm": 1.7854461669921875,
"learning_rate": 4.520995679399611e-05,
"loss": 0.20176234245300292,
"step": 830
},
{
"epoch": 1.1299052774018945,
"grad_norm": 0.9350914359092712,
"learning_rate": 4.5145266024956426e-05,
"loss": 0.1590933918952942,
"step": 835
},
{
"epoch": 1.1366711772665765,
"grad_norm": 2.1199066638946533,
"learning_rate": 4.508018828701612e-05,
"loss": 0.20157186985015868,
"step": 840
},
{
"epoch": 1.1434370771312585,
"grad_norm": 1.4717801809310913,
"learning_rate": 4.501472483024857e-05,
"loss": 0.2141291618347168,
"step": 845
},
{
"epoch": 1.1502029769959405,
"grad_norm": 1.3130544424057007,
"learning_rate": 4.494887691213639e-05,
"loss": 0.1930161952972412,
"step": 850
},
{
"epoch": 1.1569688768606226,
"grad_norm": 1.6854908466339111,
"learning_rate": 4.488264579754728e-05,
"loss": 0.16665077209472656,
"step": 855
},
{
"epoch": 1.1637347767253043,
"grad_norm": 2.862696409225464,
"learning_rate": 4.481603275870976e-05,
"loss": 0.252798056602478,
"step": 860
},
{
"epoch": 1.1705006765899864,
"grad_norm": 2.0579280853271484,
"learning_rate": 4.474903907518868e-05,
"loss": 0.22854328155517578,
"step": 865
},
{
"epoch": 1.1772665764546684,
"grad_norm": 3.0148253440856934,
"learning_rate": 4.4681666033860653e-05,
"loss": 0.18968321084976197,
"step": 870
},
{
"epoch": 1.1840324763193504,
"grad_norm": 1.6965737342834473,
"learning_rate": 4.4613914928889386e-05,
"loss": 0.18576133251190186,
"step": 875
},
{
"epoch": 1.1907983761840324,
"grad_norm": 2.1435391902923584,
"learning_rate": 4.454578706170075e-05,
"loss": 0.1876312255859375,
"step": 880
},
{
"epoch": 1.1975642760487144,
"grad_norm": 0.9975637197494507,
"learning_rate": 4.447728374095783e-05,
"loss": 0.2196826934814453,
"step": 885
},
{
"epoch": 1.2043301759133964,
"grad_norm": 2.427738666534424,
"learning_rate": 4.440840628253575e-05,
"loss": 0.18356919288635254,
"step": 890
},
{
"epoch": 1.2110960757780784,
"grad_norm": 2.2400786876678467,
"learning_rate": 4.4339156009496456e-05,
"loss": 0.17457997798919678,
"step": 895
},
{
"epoch": 1.2178619756427604,
"grad_norm": 1.907609224319458,
"learning_rate": 4.426953425206322e-05,
"loss": 0.15493079423904418,
"step": 900
},
{
"epoch": 1.2246278755074425,
"grad_norm": 1.1212676763534546,
"learning_rate": 4.4199542347595146e-05,
"loss": 0.17416157722473144,
"step": 905
},
{
"epoch": 1.2313937753721245,
"grad_norm": 2.3016531467437744,
"learning_rate": 4.4129181640561476e-05,
"loss": 0.21945109367370605,
"step": 910
},
{
"epoch": 1.2381596752368065,
"grad_norm": 2.5128695964813232,
"learning_rate": 4.405845348251573e-05,
"loss": 0.2654566287994385,
"step": 915
},
{
"epoch": 1.2449255751014885,
"grad_norm": 2.190079927444458,
"learning_rate": 4.398735923206979e-05,
"loss": 0.2536561727523804,
"step": 920
},
{
"epoch": 1.2516914749661705,
"grad_norm": 1.8736330270767212,
"learning_rate": 4.391590025486775e-05,
"loss": 0.19978296756744385,
"step": 925
},
{
"epoch": 1.2584573748308525,
"grad_norm": 1.8011804819107056,
"learning_rate": 4.384407792355976e-05,
"loss": 0.17561075687408448,
"step": 930
},
{
"epoch": 1.2652232746955345,
"grad_norm": 2.57633376121521,
"learning_rate": 4.377189361777555e-05,
"loss": 0.18484158515930177,
"step": 935
},
{
"epoch": 1.2719891745602165,
"grad_norm": 1.8354359865188599,
"learning_rate": 4.369934872409804e-05,
"loss": 0.19241479635238648,
"step": 940
},
{
"epoch": 1.2787550744248986,
"grad_norm": 2.5690648555755615,
"learning_rate": 4.3626444636036654e-05,
"loss": 0.18573579788208008,
"step": 945
},
{
"epoch": 1.2855209742895806,
"grad_norm": 2.4654386043548584,
"learning_rate": 4.355318275400051e-05,
"loss": 0.22659449577331542,
"step": 950
},
{
"epoch": 1.2922868741542626,
"grad_norm": 1.128698468208313,
"learning_rate": 4.3479564485271604e-05,
"loss": 0.22224926948547363,
"step": 955
},
{
"epoch": 1.2990527740189446,
"grad_norm": 1.55394446849823,
"learning_rate": 4.3405591243977736e-05,
"loss": 0.21414153575897216,
"step": 960
},
{
"epoch": 1.3058186738836266,
"grad_norm": 2.1317620277404785,
"learning_rate": 4.3331264451065324e-05,
"loss": 0.23089933395385742,
"step": 965
},
{
"epoch": 1.3125845737483086,
"grad_norm": 3.075485944747925,
"learning_rate": 4.325658553427215e-05,
"loss": 0.16716535091400148,
"step": 970
},
{
"epoch": 1.3193504736129906,
"grad_norm": 1.3878815174102783,
"learning_rate": 4.318155592809991e-05,
"loss": 0.20675935745239257,
"step": 975
},
{
"epoch": 1.3261163734776726,
"grad_norm": 2.4011075496673584,
"learning_rate": 4.310617707378668e-05,
"loss": 0.1940375566482544,
"step": 980
},
{
"epoch": 1.3328822733423547,
"grad_norm": 1.7323752641677856,
"learning_rate": 4.303045041927919e-05,
"loss": 0.16122257709503174,
"step": 985
},
{
"epoch": 1.3396481732070367,
"grad_norm": 3.7211077213287354,
"learning_rate": 4.2954377419205056e-05,
"loss": 0.21055731773376465,
"step": 990
},
{
"epoch": 1.3464140730717187,
"grad_norm": 1.638248324394226,
"learning_rate": 4.2877959534844826e-05,
"loss": 0.22009711265563964,
"step": 995
},
{
"epoch": 1.3531799729364005,
"grad_norm": 1.7128612995147705,
"learning_rate": 4.280119823410389e-05,
"loss": 0.20077388286590575,
"step": 1000
},
{
"epoch": 1.3599458728010825,
"grad_norm": 2.1936678886413574,
"learning_rate": 4.27240949914843e-05,
"loss": 0.18262946605682373,
"step": 1005
},
{
"epoch": 1.3667117726657645,
"grad_norm": 1.7075120210647583,
"learning_rate": 4.264665128805646e-05,
"loss": 0.18258364200592042,
"step": 1010
},
{
"epoch": 1.3734776725304465,
"grad_norm": 1.8000565767288208,
"learning_rate": 4.2568868611430636e-05,
"loss": 0.1829495310783386,
"step": 1015
},
{
"epoch": 1.3802435723951285,
"grad_norm": 1.2302501201629639,
"learning_rate": 4.249074845572842e-05,
"loss": 0.22556328773498535,
"step": 1020
},
{
"epoch": 1.3870094722598105,
"grad_norm": 2.0998246669769287,
"learning_rate": 4.2412292321554e-05,
"loss": 0.27017066478729246,
"step": 1025
},
{
"epoch": 1.3937753721244925,
"grad_norm": 2.586259365081787,
"learning_rate": 4.2333501715965386e-05,
"loss": 0.19653122425079345,
"step": 1030
},
{
"epoch": 1.4005412719891746,
"grad_norm": 2.3513941764831543,
"learning_rate": 4.2254378152445364e-05,
"loss": 0.17768259048461915,
"step": 1035
},
{
"epoch": 1.4073071718538566,
"grad_norm": 1.1629408597946167,
"learning_rate": 4.2174923150872544e-05,
"loss": 0.18216381072998047,
"step": 1040
},
{
"epoch": 1.4140730717185386,
"grad_norm": 1.9408811330795288,
"learning_rate": 4.209513823749206e-05,
"loss": 0.1798466682434082,
"step": 1045
},
{
"epoch": 1.4208389715832206,
"grad_norm": 0.8173491358757019,
"learning_rate": 4.201502494488633e-05,
"loss": 0.14073249101638793,
"step": 1050
},
{
"epoch": 1.4276048714479026,
"grad_norm": 2.334101915359497,
"learning_rate": 4.1934584811945556e-05,
"loss": 0.1975398302078247,
"step": 1055
},
{
"epoch": 1.4343707713125846,
"grad_norm": 2.1871824264526367,
"learning_rate": 4.185381938383821e-05,
"loss": 0.17156534194946288,
"step": 1060
},
{
"epoch": 1.4411366711772666,
"grad_norm": 1.924270510673523,
"learning_rate": 4.1772730211981335e-05,
"loss": 0.17501533031463623,
"step": 1065
},
{
"epoch": 1.4479025710419486,
"grad_norm": 1.500259518623352,
"learning_rate": 4.1691318854010724e-05,
"loss": 0.223386812210083,
"step": 1070
},
{
"epoch": 1.4546684709066307,
"grad_norm": 2.201751470565796,
"learning_rate": 4.160958687375104e-05,
"loss": 0.23405585289001465,
"step": 1075
},
{
"epoch": 1.4614343707713127,
"grad_norm": 2.2572615146636963,
"learning_rate": 4.152753584118576e-05,
"loss": 0.22280921936035156,
"step": 1080
},
{
"epoch": 1.4682002706359945,
"grad_norm": 2.298200845718384,
"learning_rate": 4.144516733242697e-05,
"loss": 0.21522672176361085,
"step": 1085
},
{
"epoch": 1.4749661705006765,
"grad_norm": 2.0385775566101074,
"learning_rate": 4.136248292968519e-05,
"loss": 0.20268840789794923,
"step": 1090
},
{
"epoch": 1.4817320703653585,
"grad_norm": 1.418489933013916,
"learning_rate": 4.127948422123887e-05,
"loss": 0.15638396739959717,
"step": 1095
},
{
"epoch": 1.4884979702300405,
"grad_norm": 2.208134889602661,
"learning_rate": 4.1196172801404e-05,
"loss": 0.2338188648223877,
"step": 1100
},
{
"epoch": 1.4952638700947225,
"grad_norm": 2.372204303741455,
"learning_rate": 4.111255027050334e-05,
"loss": 0.24258069992065429,
"step": 1105
},
{
"epoch": 1.5020297699594045,
"grad_norm": 2.4909098148345947,
"learning_rate": 4.1028618234835824e-05,
"loss": 0.18836082220077516,
"step": 1110
},
{
"epoch": 1.5020297699594045,
"eval_accuracy": 0.8700357104209501,
"eval_loss": 0.3137931823730469,
"eval_runtime": 29.2389,
"eval_samples_per_second": 70.044,
"eval_steps_per_second": 4.378,
"step": 1110
},
{
"epoch": 1.5087956698240865,
"grad_norm": 2.1532695293426514,
"learning_rate": 4.0944378306645606e-05,
"loss": 0.17781510353088378,
"step": 1115
},
{
"epoch": 1.5155615696887685,
"grad_norm": 3.3554134368896484,
"learning_rate": 4.085983210409114e-05,
"loss": 0.19773557186126708,
"step": 1120
},
{
"epoch": 1.5223274695534506,
"grad_norm": 1.8966747522354126,
"learning_rate": 4.0774981251214065e-05,
"loss": 0.1753719687461853,
"step": 1125
},
{
"epoch": 1.5290933694181326,
"grad_norm": 1.9342551231384277,
"learning_rate": 4.068982737790805e-05,
"loss": 0.20356435775756837,
"step": 1130
},
{
"epoch": 1.5358592692828146,
"grad_norm": 2.120213508605957,
"learning_rate": 4.0604372119887434e-05,
"loss": 0.20176234245300292,
"step": 1135
},
{
"epoch": 1.5426251691474966,
"grad_norm": 1.0722355842590332,
"learning_rate": 4.0518617118655847e-05,
"loss": 0.1417493462562561,
"step": 1140
},
{
"epoch": 1.5493910690121786,
"grad_norm": 2.388068675994873,
"learning_rate": 4.043256402147466e-05,
"loss": 0.25748581886291505,
"step": 1145
},
{
"epoch": 1.5561569688768606,
"grad_norm": 2.2098467350006104,
"learning_rate": 4.0346214481331346e-05,
"loss": 0.18525443077087403,
"step": 1150
},
{
"epoch": 1.5629228687415426,
"grad_norm": 1.7078344821929932,
"learning_rate": 4.025957015690774e-05,
"loss": 0.18819081783294678,
"step": 1155
},
{
"epoch": 1.5696887686062246,
"grad_norm": 2.038097381591797,
"learning_rate": 4.017263271254814e-05,
"loss": 0.1858762502670288,
"step": 1160
},
{
"epoch": 1.5764546684709067,
"grad_norm": 1.8393099308013916,
"learning_rate": 4.00854038182274e-05,
"loss": 0.21230158805847169,
"step": 1165
},
{
"epoch": 1.5832205683355887,
"grad_norm": 2.127931833267212,
"learning_rate": 3.9997885149518775e-05,
"loss": 0.16981457471847533,
"step": 1170
},
{
"epoch": 1.5899864682002707,
"grad_norm": 1.994484782218933,
"learning_rate": 3.99100783875618e-05,
"loss": 0.2509755611419678,
"step": 1175
},
{
"epoch": 1.5967523680649527,
"grad_norm": 1.6432493925094604,
"learning_rate": 3.9821985219029965e-05,
"loss": 0.20791516304016114,
"step": 1180
},
{
"epoch": 1.6035182679296347,
"grad_norm": 0.7421208024024963,
"learning_rate": 3.9733607336098324e-05,
"loss": 0.2141780138015747,
"step": 1185
},
{
"epoch": 1.6102841677943167,
"grad_norm": 1.1214805841445923,
"learning_rate": 3.9644946436410974e-05,
"loss": 0.17714885473251343,
"step": 1190
},
{
"epoch": 1.6170500676589987,
"grad_norm": 1.7955119609832764,
"learning_rate": 3.9556004223048484e-05,
"loss": 0.20593461990356446,
"step": 1195
},
{
"epoch": 1.6238159675236807,
"grad_norm": 1.3066389560699463,
"learning_rate": 3.946678240449515e-05,
"loss": 0.1463393449783325,
"step": 1200
},
{
"epoch": 1.6305818673883627,
"grad_norm": 1.8398350477218628,
"learning_rate": 3.937728269460616e-05,
"loss": 0.21517591476440429,
"step": 1205
},
{
"epoch": 1.6373477672530448,
"grad_norm": 1.6982916593551636,
"learning_rate": 3.928750681257473e-05,
"loss": 0.21596252918243408,
"step": 1210
},
{
"epoch": 1.6441136671177268,
"grad_norm": 1.733951449394226,
"learning_rate": 3.919745648289901e-05,
"loss": 0.21697282791137695,
"step": 1215
},
{
"epoch": 1.6508795669824088,
"grad_norm": 1.9840619564056396,
"learning_rate": 3.9107133435349016e-05,
"loss": 0.2218961238861084,
"step": 1220
},
{
"epoch": 1.6576454668470908,
"grad_norm": 2.153996706008911,
"learning_rate": 3.9016539404933406e-05,
"loss": 0.1637297034263611,
"step": 1225
},
{
"epoch": 1.6644113667117728,
"grad_norm": 1.8474509716033936,
"learning_rate": 3.892567613186606e-05,
"loss": 0.2437319278717041,
"step": 1230
},
{
"epoch": 1.6711772665764548,
"grad_norm": 1.5132358074188232,
"learning_rate": 3.88345453615328e-05,
"loss": 0.18087722063064576,
"step": 1235
},
{
"epoch": 1.6779431664411368,
"grad_norm": 1.9187140464782715,
"learning_rate": 3.874314884445771e-05,
"loss": 0.1810509204864502,
"step": 1240
},
{
"epoch": 1.6847090663058186,
"grad_norm": 2.858391046524048,
"learning_rate": 3.865148833626963e-05,
"loss": 0.2332247257232666,
"step": 1245
},
{
"epoch": 1.6914749661705006,
"grad_norm": 2.2533044815063477,
"learning_rate": 3.855956559766836e-05,
"loss": 0.19411702156066896,
"step": 1250
},
{
"epoch": 1.6982408660351827,
"grad_norm": 1.7951242923736572,
"learning_rate": 3.8467382394390895e-05,
"loss": 0.1899639844894409,
"step": 1255
},
{
"epoch": 1.7050067658998647,
"grad_norm": 1.1991900205612183,
"learning_rate": 3.8374940497177434e-05,
"loss": 0.182519793510437,
"step": 1260
},
{
"epoch": 1.7117726657645467,
"grad_norm": 1.862929344177246,
"learning_rate": 3.8282241681737445e-05,
"loss": 0.24091653823852538,
"step": 1265
},
{
"epoch": 1.7185385656292287,
"grad_norm": 1.7019075155258179,
"learning_rate": 3.8189287728715504e-05,
"loss": 0.2029292106628418,
"step": 1270
},
{
"epoch": 1.7253044654939107,
"grad_norm": 2.528502941131592,
"learning_rate": 3.809608042365711e-05,
"loss": 0.16294546127319337,
"step": 1275
},
{
"epoch": 1.7320703653585927,
"grad_norm": 1.1748173236846924,
"learning_rate": 3.8002621556974367e-05,
"loss": 0.20352749824523925,
"step": 1280
},
{
"epoch": 1.7388362652232747,
"grad_norm": 0.7170626521110535,
"learning_rate": 3.790891292391163e-05,
"loss": 0.14799299240112304,
"step": 1285
},
{
"epoch": 1.7456021650879567,
"grad_norm": 1.1239309310913086,
"learning_rate": 3.781495632451098e-05,
"loss": 0.20588746070861816,
"step": 1290
},
{
"epoch": 1.7523680649526387,
"grad_norm": 1.6925621032714844,
"learning_rate": 3.772075356357768e-05,
"loss": 0.1684184193611145,
"step": 1295
},
{
"epoch": 1.7591339648173205,
"grad_norm": 1.6986865997314453,
"learning_rate": 3.7626306450645474e-05,
"loss": 0.17929956912994385,
"step": 1300
},
{
"epoch": 1.7658998646820026,
"grad_norm": 1.2348734140396118,
"learning_rate": 3.753161679994186e-05,
"loss": 0.1621527910232544,
"step": 1305
},
{
"epoch": 1.7726657645466846,
"grad_norm": 2.7075700759887695,
"learning_rate": 3.743668643035321e-05,
"loss": 0.2273587703704834,
"step": 1310
},
{
"epoch": 1.7794316644113666,
"grad_norm": 1.9679162502288818,
"learning_rate": 3.734151716538987e-05,
"loss": 0.17612202167510987,
"step": 1315
},
{
"epoch": 1.7861975642760486,
"grad_norm": 1.1956671476364136,
"learning_rate": 3.724611083315109e-05,
"loss": 0.16846781969070435,
"step": 1320
},
{
"epoch": 1.7929634641407306,
"grad_norm": 1.5890071392059326,
"learning_rate": 3.715046926628994e-05,
"loss": 0.18916544914245606,
"step": 1325
},
{
"epoch": 1.7997293640054126,
"grad_norm": 1.2714638710021973,
"learning_rate": 3.7054594301978074e-05,
"loss": 0.16342637538909913,
"step": 1330
},
{
"epoch": 1.8064952638700946,
"grad_norm": 1.6062065362930298,
"learning_rate": 3.6958487781870486e-05,
"loss": 0.21015429496765137,
"step": 1335
},
{
"epoch": 1.8132611637347766,
"grad_norm": 2.673487424850464,
"learning_rate": 3.686215155207008e-05,
"loss": 0.16505404710769653,
"step": 1340
},
{
"epoch": 1.8200270635994586,
"grad_norm": 1.8138383626937866,
"learning_rate": 3.6765587463092276e-05,
"loss": 0.21825101375579833,
"step": 1345
},
{
"epoch": 1.8267929634641407,
"grad_norm": 2.180575370788574,
"learning_rate": 3.6668797369829386e-05,
"loss": 0.2074803113937378,
"step": 1350
},
{
"epoch": 1.8335588633288227,
"grad_norm": 1.7296091318130493,
"learning_rate": 3.657178313151504e-05,
"loss": 0.19976370334625243,
"step": 1355
},
{
"epoch": 1.8403247631935047,
"grad_norm": 1.6362652778625488,
"learning_rate": 3.6474546611688445e-05,
"loss": 0.1672407269477844,
"step": 1360
},
{
"epoch": 1.8470906630581867,
"grad_norm": 1.4946094751358032,
"learning_rate": 3.63770896781586e-05,
"loss": 0.13958334922790527,
"step": 1365
},
{
"epoch": 1.8538565629228687,
"grad_norm": 2.1709749698638916,
"learning_rate": 3.627941420296841e-05,
"loss": 0.22573065757751465,
"step": 1370
},
{
"epoch": 1.8606224627875507,
"grad_norm": 1.0254132747650146,
"learning_rate": 3.6181522062358755e-05,
"loss": 0.14214818477630614,
"step": 1375
},
{
"epoch": 1.8673883626522327,
"grad_norm": 2.482468366622925,
"learning_rate": 3.608341513673238e-05,
"loss": 0.1950739622116089,
"step": 1380
},
{
"epoch": 1.8741542625169147,
"grad_norm": 3.2865729331970215,
"learning_rate": 3.5985095310617846e-05,
"loss": 0.22073097229003907,
"step": 1385
},
{
"epoch": 1.8809201623815968,
"grad_norm": 2.1295671463012695,
"learning_rate": 3.5886564472633323e-05,
"loss": 0.17798930406570435,
"step": 1390
},
{
"epoch": 1.8876860622462788,
"grad_norm": 1.7020217180252075,
"learning_rate": 3.5787824515450254e-05,
"loss": 0.23192262649536133,
"step": 1395
},
{
"epoch": 1.8944519621109608,
"grad_norm": 2.24668550491333,
"learning_rate": 3.568887733575706e-05,
"loss": 0.18974560499191284,
"step": 1400
},
{
"epoch": 1.9012178619756428,
"grad_norm": 1.2608585357666016,
"learning_rate": 3.5589724834222666e-05,
"loss": 0.17455393075942993,
"step": 1405
},
{
"epoch": 1.9079837618403248,
"grad_norm": 1.7468396425247192,
"learning_rate": 3.549036891546004e-05,
"loss": 0.15837132930755615,
"step": 1410
},
{
"epoch": 1.9147496617050068,
"grad_norm": 2.787322521209717,
"learning_rate": 3.539081148798954e-05,
"loss": 0.21066501140594482,
"step": 1415
},
{
"epoch": 1.9215155615696888,
"grad_norm": 2.0656638145446777,
"learning_rate": 3.5291054464202304e-05,
"loss": 0.17975693941116333,
"step": 1420
},
{
"epoch": 1.9282814614343708,
"grad_norm": 1.377244234085083,
"learning_rate": 3.519109976032351e-05,
"loss": 0.20913619995117189,
"step": 1425
},
{
"epoch": 1.9350473612990529,
"grad_norm": 2.588541030883789,
"learning_rate": 3.509094929637556e-05,
"loss": 0.20017833709716798,
"step": 1430
},
{
"epoch": 1.9418132611637349,
"grad_norm": 1.9158939123153687,
"learning_rate": 3.499060499614117e-05,
"loss": 0.2471320152282715,
"step": 1435
},
{
"epoch": 1.9485791610284169,
"grad_norm": 3.4396183490753174,
"learning_rate": 3.489006878712647e-05,
"loss": 0.188225257396698,
"step": 1440
},
{
"epoch": 1.955345060893099,
"grad_norm": 2.0203394889831543,
"learning_rate": 3.478934260052394e-05,
"loss": 0.21033647060394287,
"step": 1445
},
{
"epoch": 1.962110960757781,
"grad_norm": 1.9122129678726196,
"learning_rate": 3.468842837117532e-05,
"loss": 0.15714671611785888,
"step": 1450
},
{
"epoch": 1.968876860622463,
"grad_norm": 1.2576748132705688,
"learning_rate": 3.458732803753446e-05,
"loss": 0.17848788499832152,
"step": 1455
},
{
"epoch": 1.975642760487145,
"grad_norm": 3.0754387378692627,
"learning_rate": 3.448604354163007e-05,
"loss": 0.1968851327896118,
"step": 1460
},
{
"epoch": 1.982408660351827,
"grad_norm": 1.8255152702331543,
"learning_rate": 3.438457682902841e-05,
"loss": 0.17655500173568725,
"step": 1465
},
{
"epoch": 1.989174560216509,
"grad_norm": 1.609027624130249,
"learning_rate": 3.4282929848795944e-05,
"loss": 0.1948060154914856,
"step": 1470
},
{
"epoch": 1.9959404600811907,
"grad_norm": 2.263242721557617,
"learning_rate": 3.418110455346186e-05,
"loss": 0.22324516773223876,
"step": 1475
},
{
"epoch": 2.002706359945873,
"grad_norm": 1.7252802848815918,
"learning_rate": 3.4079102898980633e-05,
"loss": 0.19495820999145508,
"step": 1480
},
{
"epoch": 2.002706359945873,
"eval_accuracy": 0.869602856833676,
"eval_loss": 0.3007725179195404,
"eval_runtime": 29.1072,
"eval_samples_per_second": 70.361,
"eval_steps_per_second": 4.398,
"step": 1480
},
{
"epoch": 2.009472259810555,
"grad_norm": 1.6615360975265503,
"learning_rate": 3.397692684469434e-05,
"loss": 0.1471646785736084,
"step": 1485
},
{
"epoch": 2.016238159675237,
"grad_norm": 1.3666348457336426,
"learning_rate": 3.387457835329513e-05,
"loss": 0.20376362800598144,
"step": 1490
},
{
"epoch": 2.023004059539919,
"grad_norm": 2.3382604122161865,
"learning_rate": 3.3772059390787496e-05,
"loss": 0.22116050720214844,
"step": 1495
},
{
"epoch": 2.029769959404601,
"grad_norm": 2.0877110958099365,
"learning_rate": 3.3669371926450486e-05,
"loss": 0.14569587707519532,
"step": 1500
},
{
"epoch": 2.0365358592692826,
"grad_norm": 1.2946667671203613,
"learning_rate": 3.356651793279986e-05,
"loss": 0.1713175058364868,
"step": 1505
},
{
"epoch": 2.0433017591339646,
"grad_norm": 2.0007376670837402,
"learning_rate": 3.346349938555027e-05,
"loss": 0.13855953216552735,
"step": 1510
},
{
"epoch": 2.0500676589986466,
"grad_norm": 2.054121732711792,
"learning_rate": 3.336031826357724e-05,
"loss": 0.16211364269256592,
"step": 1515
},
{
"epoch": 2.0568335588633286,
"grad_norm": 2.056433916091919,
"learning_rate": 3.3256976548879184e-05,
"loss": 0.20744428634643555,
"step": 1520
},
{
"epoch": 2.0635994587280106,
"grad_norm": 1.693894863128662,
"learning_rate": 3.315347622653932e-05,
"loss": 0.20456321239471437,
"step": 1525
},
{
"epoch": 2.0703653585926927,
"grad_norm": 1.5622609853744507,
"learning_rate": 3.304981928468758e-05,
"loss": 0.19015119075775147,
"step": 1530
},
{
"epoch": 2.0771312584573747,
"grad_norm": 2.1416757106781006,
"learning_rate": 3.2946007714462355e-05,
"loss": 0.19555115699768066,
"step": 1535
},
{
"epoch": 2.0838971583220567,
"grad_norm": 1.2763415575027466,
"learning_rate": 3.284204350997229e-05,
"loss": 0.19667301177978516,
"step": 1540
},
{
"epoch": 2.0906630581867387,
"grad_norm": 1.879966139793396,
"learning_rate": 3.273792866825798e-05,
"loss": 0.13344597816467285,
"step": 1545
},
{
"epoch": 2.0974289580514207,
"grad_norm": 1.442071557044983,
"learning_rate": 3.26336651892536e-05,
"loss": 0.18474416732788085,
"step": 1550
},
{
"epoch": 2.1041948579161027,
"grad_norm": 2.347238540649414,
"learning_rate": 3.252925507574846e-05,
"loss": 0.18224793672561646,
"step": 1555
},
{
"epoch": 2.1109607577807847,
"grad_norm": 1.4569040536880493,
"learning_rate": 3.242470033334859e-05,
"loss": 0.16616876125335694,
"step": 1560
},
{
"epoch": 2.1177266576454667,
"grad_norm": 1.9158858060836792,
"learning_rate": 3.232000297043818e-05,
"loss": 0.17404348850250245,
"step": 1565
},
{
"epoch": 2.1244925575101488,
"grad_norm": 2.5596396923065186,
"learning_rate": 3.2215164998141014e-05,
"loss": 0.19700939655303956,
"step": 1570
},
{
"epoch": 2.1312584573748308,
"grad_norm": 1.3152157068252563,
"learning_rate": 3.21101884302818e-05,
"loss": 0.13300225734710694,
"step": 1575
},
{
"epoch": 2.138024357239513,
"grad_norm": 2.5467565059661865,
"learning_rate": 3.200507528334754e-05,
"loss": 0.1695353150367737,
"step": 1580
},
{
"epoch": 2.144790257104195,
"grad_norm": 1.9991508722305298,
"learning_rate": 3.189982757644879e-05,
"loss": 0.1894860029220581,
"step": 1585
},
{
"epoch": 2.151556156968877,
"grad_norm": 1.9280539751052856,
"learning_rate": 3.179444733128083e-05,
"loss": 0.18253859281539916,
"step": 1590
},
{
"epoch": 2.158322056833559,
"grad_norm": 1.918892741203308,
"learning_rate": 3.168893657208485e-05,
"loss": 0.18120259046554565,
"step": 1595
},
{
"epoch": 2.165087956698241,
"grad_norm": 1.3024680614471436,
"learning_rate": 3.158329732560912e-05,
"loss": 0.1363304615020752,
"step": 1600
},
{
"epoch": 2.171853856562923,
"grad_norm": 1.9482077360153198,
"learning_rate": 3.147753162106997e-05,
"loss": 0.17441680431365966,
"step": 1605
},
{
"epoch": 2.178619756427605,
"grad_norm": 2.75502347946167,
"learning_rate": 3.137164149011287e-05,
"loss": 0.18927013874053955,
"step": 1610
},
{
"epoch": 2.185385656292287,
"grad_norm": 1.0601168870925903,
"learning_rate": 3.126562896677338e-05,
"loss": 0.1616207480430603,
"step": 1615
},
{
"epoch": 2.192151556156969,
"grad_norm": 2.5636837482452393,
"learning_rate": 3.1159496087438094e-05,
"loss": 0.1545897126197815,
"step": 1620
},
{
"epoch": 2.198917456021651,
"grad_norm": 1.3877092599868774,
"learning_rate": 3.105324489080551e-05,
"loss": 0.17156057357788085,
"step": 1625
},
{
"epoch": 2.205683355886333,
"grad_norm": 1.7555543184280396,
"learning_rate": 3.094687741784688e-05,
"loss": 0.16101455688476562,
"step": 1630
},
{
"epoch": 2.212449255751015,
"grad_norm": 1.5731257200241089,
"learning_rate": 3.0840395711767e-05,
"loss": 0.1715109944343567,
"step": 1635
},
{
"epoch": 2.219215155615697,
"grad_norm": 1.6472532749176025,
"learning_rate": 3.073380181796495e-05,
"loss": 0.16718752384185792,
"step": 1640
},
{
"epoch": 2.225981055480379,
"grad_norm": 1.1000165939331055,
"learning_rate": 3.062709778399482e-05,
"loss": 0.14798297882080078,
"step": 1645
},
{
"epoch": 2.232746955345061,
"grad_norm": 1.2627995014190674,
"learning_rate": 3.052028565952639e-05,
"loss": 0.15804848670959473,
"step": 1650
},
{
"epoch": 2.239512855209743,
"grad_norm": 1.319171667098999,
"learning_rate": 3.0413367496305724e-05,
"loss": 0.16094930171966554,
"step": 1655
},
{
"epoch": 2.246278755074425,
"grad_norm": 1.750583529472351,
"learning_rate": 3.0306345348115767e-05,
"loss": 0.19895588159561156,
"step": 1660
},
{
"epoch": 2.253044654939107,
"grad_norm": 1.338939905166626,
"learning_rate": 3.019922127073694e-05,
"loss": 0.16100188493728637,
"step": 1665
},
{
"epoch": 2.259810554803789,
"grad_norm": 2.591289520263672,
"learning_rate": 3.009199732190757e-05,
"loss": 0.1722928524017334,
"step": 1670
},
{
"epoch": 2.266576454668471,
"grad_norm": 1.9176908731460571,
"learning_rate": 2.9984675561284457e-05,
"loss": 0.13343515396118164,
"step": 1675
},
{
"epoch": 2.273342354533153,
"grad_norm": 1.6658287048339844,
"learning_rate": 2.9877258050403212e-05,
"loss": 0.162172269821167,
"step": 1680
},
{
"epoch": 2.280108254397835,
"grad_norm": 1.417536973953247,
"learning_rate": 2.9769746852638726e-05,
"loss": 0.16868025064468384,
"step": 1685
},
{
"epoch": 2.286874154262517,
"grad_norm": 1.3248891830444336,
"learning_rate": 2.966214403316553e-05,
"loss": 0.18633471727371215,
"step": 1690
},
{
"epoch": 2.293640054127199,
"grad_norm": 2.6158204078674316,
"learning_rate": 2.9554451658918076e-05,
"loss": 0.21243324279785156,
"step": 1695
},
{
"epoch": 2.300405953991881,
"grad_norm": 2.3820583820343018,
"learning_rate": 2.944667179855109e-05,
"loss": 0.20705819129943848,
"step": 1700
},
{
"epoch": 2.307171853856563,
"grad_norm": 1.748436450958252,
"learning_rate": 2.9338806522399816e-05,
"loss": 0.13948395252227783,
"step": 1705
},
{
"epoch": 2.313937753721245,
"grad_norm": 1.296347737312317,
"learning_rate": 2.923085790244023e-05,
"loss": 0.17860794067382812,
"step": 1710
},
{
"epoch": 2.320703653585927,
"grad_norm": 1.3463002443313599,
"learning_rate": 2.9122828012249253e-05,
"loss": 0.16463401317596435,
"step": 1715
},
{
"epoch": 2.3274695534506087,
"grad_norm": 2.2493791580200195,
"learning_rate": 2.901471892696493e-05,
"loss": 0.16881377696990968,
"step": 1720
},
{
"epoch": 2.3342354533152907,
"grad_norm": 1.6002099514007568,
"learning_rate": 2.890653272324655e-05,
"loss": 0.16787372827529906,
"step": 1725
},
{
"epoch": 2.3410013531799727,
"grad_norm": 1.6656807661056519,
"learning_rate": 2.879827147923476e-05,
"loss": 0.1768743395805359,
"step": 1730
},
{
"epoch": 2.3477672530446547,
"grad_norm": 1.7007827758789062,
"learning_rate": 2.868993727451168e-05,
"loss": 0.15016405582427977,
"step": 1735
},
{
"epoch": 2.3545331529093367,
"grad_norm": 3.33925199508667,
"learning_rate": 2.858153219006088e-05,
"loss": 0.1700860619544983,
"step": 1740
},
{
"epoch": 2.3612990527740187,
"grad_norm": 1.796484112739563,
"learning_rate": 2.8473058308227495e-05,
"loss": 0.1265573263168335,
"step": 1745
},
{
"epoch": 2.3680649526387008,
"grad_norm": 1.6851625442504883,
"learning_rate": 2.8364517712678156e-05,
"loss": 0.19020633697509765,
"step": 1750
},
{
"epoch": 2.3748308525033828,
"grad_norm": 1.8832194805145264,
"learning_rate": 2.8255912488361008e-05,
"loss": 0.16289949417114258,
"step": 1755
},
{
"epoch": 2.381596752368065,
"grad_norm": 1.9859877824783325,
"learning_rate": 2.8147244721465636e-05,
"loss": 0.16884443759918213,
"step": 1760
},
{
"epoch": 2.388362652232747,
"grad_norm": 1.9740185737609863,
"learning_rate": 2.8038516499383012e-05,
"loss": 0.17444243431091308,
"step": 1765
},
{
"epoch": 2.395128552097429,
"grad_norm": 0.9951803684234619,
"learning_rate": 2.792972991066537e-05,
"loss": 0.19391849040985107,
"step": 1770
},
{
"epoch": 2.401894451962111,
"grad_norm": 1.3108264207839966,
"learning_rate": 2.7820887044986128e-05,
"loss": 0.1552678346633911,
"step": 1775
},
{
"epoch": 2.408660351826793,
"grad_norm": 2.8585104942321777,
"learning_rate": 2.77119899930997e-05,
"loss": 0.14544509649276732,
"step": 1780
},
{
"epoch": 2.415426251691475,
"grad_norm": 3.70153546333313,
"learning_rate": 2.760304084680137e-05,
"loss": 0.21861593723297118,
"step": 1785
},
{
"epoch": 2.422192151556157,
"grad_norm": 2.1083192825317383,
"learning_rate": 2.7494041698887102e-05,
"loss": 0.19854121208190917,
"step": 1790
},
{
"epoch": 2.428958051420839,
"grad_norm": 1.2471938133239746,
"learning_rate": 2.7384994643113333e-05,
"loss": 0.1913970708847046,
"step": 1795
},
{
"epoch": 2.435723951285521,
"grad_norm": 1.8073750734329224,
"learning_rate": 2.7275901774156754e-05,
"loss": 0.1438676357269287,
"step": 1800
},
{
"epoch": 2.442489851150203,
"grad_norm": 1.0980958938598633,
"learning_rate": 2.7166765187574094e-05,
"loss": 0.1278759241104126,
"step": 1805
},
{
"epoch": 2.449255751014885,
"grad_norm": 1.7975143194198608,
"learning_rate": 2.7057586979761833e-05,
"loss": 0.17225244045257568,
"step": 1810
},
{
"epoch": 2.456021650879567,
"grad_norm": 0.5247082114219666,
"learning_rate": 2.6948369247915972e-05,
"loss": 0.12868423461914064,
"step": 1815
},
{
"epoch": 2.462787550744249,
"grad_norm": 1.7603271007537842,
"learning_rate": 2.683911408999169e-05,
"loss": 0.23005156517028807,
"step": 1820
},
{
"epoch": 2.469553450608931,
"grad_norm": 2.184208393096924,
"learning_rate": 2.6729823604663108e-05,
"loss": 0.19608937501907348,
"step": 1825
},
{
"epoch": 2.476319350473613,
"grad_norm": 2.2964539527893066,
"learning_rate": 2.6620499891282952e-05,
"loss": 0.21042485237121583,
"step": 1830
},
{
"epoch": 2.483085250338295,
"grad_norm": 1.8081086874008179,
"learning_rate": 2.6511145049842208e-05,
"loss": 0.17142150402069092,
"step": 1835
},
{
"epoch": 2.489851150202977,
"grad_norm": 1.5499011278152466,
"learning_rate": 2.6401761180929797e-05,
"loss": 0.16137609481811524,
"step": 1840
},
{
"epoch": 2.496617050067659,
"grad_norm": 1.562762975692749,
"learning_rate": 2.6292350385692244e-05,
"loss": 0.20409109592437744,
"step": 1845
},
{
"epoch": 2.503382949932341,
"grad_norm": 2.439056634902954,
"learning_rate": 2.6182914765793294e-05,
"loss": 0.17071313858032228,
"step": 1850
},
{
"epoch": 2.503382949932341,
"eval_accuracy": 0.8742560328968726,
"eval_loss": 0.3011792302131653,
"eval_runtime": 28.4313,
"eval_samples_per_second": 72.033,
"eval_steps_per_second": 4.502,
"step": 1850
},
{
"epoch": 2.510148849797023,
"grad_norm": 1.732704758644104,
"learning_rate": 2.6073456423373528e-05,
"loss": 0.17599669694900513,
"step": 1855
},
{
"epoch": 2.516914749661705,
"grad_norm": 2.0267298221588135,
"learning_rate": 2.596397746101002e-05,
"loss": 0.1670513391494751,
"step": 1860
},
{
"epoch": 2.523680649526387,
"grad_norm": 1.6351135969161987,
"learning_rate": 2.5854479981675933e-05,
"loss": 0.13289605379104613,
"step": 1865
},
{
"epoch": 2.530446549391069,
"grad_norm": 2.1412038803100586,
"learning_rate": 2.5744966088700117e-05,
"loss": 0.1689595937728882,
"step": 1870
},
{
"epoch": 2.537212449255751,
"grad_norm": 1.6307116746902466,
"learning_rate": 2.563543788572671e-05,
"loss": 0.17463994026184082,
"step": 1875
},
{
"epoch": 2.543978349120433,
"grad_norm": 3.351099967956543,
"learning_rate": 2.552589747667472e-05,
"loss": 0.23298258781433107,
"step": 1880
},
{
"epoch": 2.550744248985115,
"grad_norm": 3.3022725582122803,
"learning_rate": 2.5416346965697647e-05,
"loss": 0.16928234100341796,
"step": 1885
},
{
"epoch": 2.557510148849797,
"grad_norm": 1.3672479391098022,
"learning_rate": 2.5306788457143e-05,
"loss": 0.1693511962890625,
"step": 1890
},
{
"epoch": 2.564276048714479,
"grad_norm": 2.323725938796997,
"learning_rate": 2.5197224055511943e-05,
"loss": 0.13905444145202636,
"step": 1895
},
{
"epoch": 2.571041948579161,
"grad_norm": 2.6577937602996826,
"learning_rate": 2.5087655865418835e-05,
"loss": 0.15658373832702638,
"step": 1900
},
{
"epoch": 2.577807848443843,
"grad_norm": 1.1090679168701172,
"learning_rate": 2.497808599155079e-05,
"loss": 0.14420833587646484,
"step": 1905
},
{
"epoch": 2.584573748308525,
"grad_norm": 1.3389736413955688,
"learning_rate": 2.48685165386273e-05,
"loss": 0.14889607429504395,
"step": 1910
},
{
"epoch": 2.591339648173207,
"grad_norm": 1.582906723022461,
"learning_rate": 2.4758949611359726e-05,
"loss": 0.1745772361755371,
"step": 1915
},
{
"epoch": 2.598105548037889,
"grad_norm": 1.972942590713501,
"learning_rate": 2.4649387314410945e-05,
"loss": 0.1478969097137451,
"step": 1920
},
{
"epoch": 2.604871447902571,
"grad_norm": 1.7525840997695923,
"learning_rate": 2.4539831752354884e-05,
"loss": 0.14707398414611816,
"step": 1925
},
{
"epoch": 2.611637347767253,
"grad_norm": 1.3981934785842896,
"learning_rate": 2.4430285029636095e-05,
"loss": 0.1847022294998169,
"step": 1930
},
{
"epoch": 2.618403247631935,
"grad_norm": 1.6785058975219727,
"learning_rate": 2.4320749250529347e-05,
"loss": 0.13264868259429932,
"step": 1935
},
{
"epoch": 2.6251691474966172,
"grad_norm": 2.9816620349884033,
"learning_rate": 2.4211226519099182e-05,
"loss": 0.14727123975753784,
"step": 1940
},
{
"epoch": 2.6319350473612992,
"grad_norm": 2.032968044281006,
"learning_rate": 2.410171893915952e-05,
"loss": 0.17831077575683593,
"step": 1945
},
{
"epoch": 2.6387009472259813,
"grad_norm": 1.4775835275650024,
"learning_rate": 2.3992228614233216e-05,
"loss": 0.1569720149040222,
"step": 1950
},
{
"epoch": 2.6454668470906633,
"grad_norm": 1.9534945487976074,
"learning_rate": 2.3882757647511704e-05,
"loss": 0.14980229139328002,
"step": 1955
},
{
"epoch": 2.6522327469553453,
"grad_norm": 3.1628122329711914,
"learning_rate": 2.3773308141814552e-05,
"loss": 0.21141133308410645,
"step": 1960
},
{
"epoch": 2.6589986468200273,
"grad_norm": 1.6550230979919434,
"learning_rate": 2.3663882199549083e-05,
"loss": 0.17134083509445192,
"step": 1965
},
{
"epoch": 2.6657645466847093,
"grad_norm": 1.128572940826416,
"learning_rate": 2.355448192267e-05,
"loss": 0.13878700733184815,
"step": 1970
},
{
"epoch": 2.6725304465493913,
"grad_norm": 1.9364409446716309,
"learning_rate": 2.3445109412638993e-05,
"loss": 0.23696985244750976,
"step": 1975
},
{
"epoch": 2.6792963464140733,
"grad_norm": 1.8815973997116089,
"learning_rate": 2.3335766770384402e-05,
"loss": 0.14180320501327515,
"step": 1980
},
{
"epoch": 2.6860622462787553,
"grad_norm": 1.4095399379730225,
"learning_rate": 2.3226456096260797e-05,
"loss": 0.135154128074646,
"step": 1985
},
{
"epoch": 2.6928281461434374,
"grad_norm": 2.3532071113586426,
"learning_rate": 2.3117179490008704e-05,
"loss": 0.1294294476509094,
"step": 1990
},
{
"epoch": 2.699594046008119,
"grad_norm": 1.5369846820831299,
"learning_rate": 2.300793905071425e-05,
"loss": 0.12942357063293458,
"step": 1995
},
{
"epoch": 2.706359945872801,
"grad_norm": 1.5489847660064697,
"learning_rate": 2.2898736876768815e-05,
"loss": 0.16081830263137817,
"step": 2000
},
{
"epoch": 2.713125845737483,
"grad_norm": 1.658352255821228,
"learning_rate": 2.2789575065828745e-05,
"loss": 0.1314360737800598,
"step": 2005
},
{
"epoch": 2.719891745602165,
"grad_norm": 1.6515543460845947,
"learning_rate": 2.268045571477507e-05,
"loss": 0.18171358108520508,
"step": 2010
},
{
"epoch": 2.726657645466847,
"grad_norm": 2.0657124519348145,
"learning_rate": 2.2571380919673192e-05,
"loss": 0.16487687826156616,
"step": 2015
},
{
"epoch": 2.733423545331529,
"grad_norm": 1.5443483591079712,
"learning_rate": 2.2462352775732654e-05,
"loss": 0.134649920463562,
"step": 2020
},
{
"epoch": 2.740189445196211,
"grad_norm": 1.5175132751464844,
"learning_rate": 2.2353373377266873e-05,
"loss": 0.15421329736709594,
"step": 2025
},
{
"epoch": 2.746955345060893,
"grad_norm": 1.5925195217132568,
"learning_rate": 2.2244444817652922e-05,
"loss": 0.14144442081451417,
"step": 2030
},
{
"epoch": 2.753721244925575,
"grad_norm": 1.248831033706665,
"learning_rate": 2.2135569189291322e-05,
"loss": 0.15530011653900147,
"step": 2035
},
{
"epoch": 2.760487144790257,
"grad_norm": 2.5789451599121094,
"learning_rate": 2.2026748583565824e-05,
"loss": 0.15484405755996705,
"step": 2040
},
{
"epoch": 2.767253044654939,
"grad_norm": 1.7521063089370728,
"learning_rate": 2.191798509080327e-05,
"loss": 0.16093764305114747,
"step": 2045
},
{
"epoch": 2.774018944519621,
"grad_norm": 1.2671983242034912,
"learning_rate": 2.1809280800233418e-05,
"loss": 0.17099790573120116,
"step": 2050
},
{
"epoch": 2.780784844384303,
"grad_norm": 2.7466530799865723,
"learning_rate": 2.1700637799948796e-05,
"loss": 0.19362075328826905,
"step": 2055
},
{
"epoch": 2.787550744248985,
"grad_norm": 2.1307995319366455,
"learning_rate": 2.1592058176864645e-05,
"loss": 0.17789866924285888,
"step": 2060
},
{
"epoch": 2.794316644113667,
"grad_norm": 0.9145678877830505,
"learning_rate": 2.1483544016678772e-05,
"loss": 0.17511709928512573,
"step": 2065
},
{
"epoch": 2.801082543978349,
"grad_norm": 1.640244722366333,
"learning_rate": 2.1375097403831533e-05,
"loss": 0.14006191492080688,
"step": 2070
},
{
"epoch": 2.807848443843031,
"grad_norm": 2.0589239597320557,
"learning_rate": 2.1266720421465755e-05,
"loss": 0.19432661533355713,
"step": 2075
},
{
"epoch": 2.814614343707713,
"grad_norm": 1.4617047309875488,
"learning_rate": 2.1158415151386744e-05,
"loss": 0.17739524841308593,
"step": 2080
},
{
"epoch": 2.821380243572395,
"grad_norm": 1.34241783618927,
"learning_rate": 2.1050183674022304e-05,
"loss": 0.1305454730987549,
"step": 2085
},
{
"epoch": 2.828146143437077,
"grad_norm": 2.4104013442993164,
"learning_rate": 2.094202806838273e-05,
"loss": 0.12868118286132812,
"step": 2090
},
{
"epoch": 2.834912043301759,
"grad_norm": 2.0952541828155518,
"learning_rate": 2.083395041202093e-05,
"loss": 0.19703412055969238,
"step": 2095
},
{
"epoch": 2.841677943166441,
"grad_norm": 2.0270049571990967,
"learning_rate": 2.0725952780992468e-05,
"loss": 0.11001923084259033,
"step": 2100
},
{
"epoch": 2.848443843031123,
"grad_norm": 1.680545449256897,
"learning_rate": 2.0618037249815735e-05,
"loss": 0.16156392097473143,
"step": 2105
},
{
"epoch": 2.855209742895805,
"grad_norm": 1.8292032480239868,
"learning_rate": 2.051020589143205e-05,
"loss": 0.16260061264038086,
"step": 2110
},
{
"epoch": 2.861975642760487,
"grad_norm": 2.2971911430358887,
"learning_rate": 2.040246077716587e-05,
"loss": 0.20823984146118163,
"step": 2115
},
{
"epoch": 2.8687415426251692,
"grad_norm": 1.5021730661392212,
"learning_rate": 2.0294803976685003e-05,
"loss": 0.19497815370559693,
"step": 2120
},
{
"epoch": 2.8755074424898512,
"grad_norm": 1.9159224033355713,
"learning_rate": 2.018723755796082e-05,
"loss": 0.20413491725921631,
"step": 2125
},
{
"epoch": 2.8822733423545333,
"grad_norm": 2.2621889114379883,
"learning_rate": 2.007976358722858e-05,
"loss": 0.15241941213607788,
"step": 2130
},
{
"epoch": 2.8890392422192153,
"grad_norm": 2.4139363765716553,
"learning_rate": 1.9972384128947703e-05,
"loss": 0.16838457584381103,
"step": 2135
},
{
"epoch": 2.8958051420838973,
"grad_norm": 1.559711217880249,
"learning_rate": 1.986510124576214e-05,
"loss": 0.12026944160461425,
"step": 2140
},
{
"epoch": 2.9025710419485793,
"grad_norm": 2.1625890731811523,
"learning_rate": 1.9757916998460723e-05,
"loss": 0.1890716314315796,
"step": 2145
},
{
"epoch": 2.9093369418132613,
"grad_norm": 1.5559502840042114,
"learning_rate": 1.9650833445937604e-05,
"loss": 0.146738600730896,
"step": 2150
},
{
"epoch": 2.9161028416779433,
"grad_norm": 0.9184536933898926,
"learning_rate": 1.9543852645152693e-05,
"loss": 0.13018094301223754,
"step": 2155
},
{
"epoch": 2.9228687415426253,
"grad_norm": 1.670619010925293,
"learning_rate": 1.9436976651092144e-05,
"loss": 0.1535303235054016,
"step": 2160
},
{
"epoch": 2.9296346414073073,
"grad_norm": 3.0414466857910156,
"learning_rate": 1.9330207516728894e-05,
"loss": 0.15447769165039063,
"step": 2165
},
{
"epoch": 2.936400541271989,
"grad_norm": 2.4599900245666504,
"learning_rate": 1.9223547292983223e-05,
"loss": 0.1688704490661621,
"step": 2170
},
{
"epoch": 2.943166441136671,
"grad_norm": 1.2857577800750732,
"learning_rate": 1.9116998028683346e-05,
"loss": 0.18663138151168823,
"step": 2175
},
{
"epoch": 2.949932341001353,
"grad_norm": 2.0454773902893066,
"learning_rate": 1.9010561770526076e-05,
"loss": 0.17161012887954713,
"step": 2180
},
{
"epoch": 2.956698240866035,
"grad_norm": 1.822501540184021,
"learning_rate": 1.8904240563037488e-05,
"loss": 0.20146393775939941,
"step": 2185
},
{
"epoch": 2.963464140730717,
"grad_norm": 1.5670403242111206,
"learning_rate": 1.879803644853368e-05,
"loss": 0.11614130735397339,
"step": 2190
},
{
"epoch": 2.970230040595399,
"grad_norm": 1.9222493171691895,
"learning_rate": 1.8691951467081485e-05,
"loss": 0.2150808811187744,
"step": 2195
},
{
"epoch": 2.976995940460081,
"grad_norm": 3.534147262573242,
"learning_rate": 1.8585987656459336e-05,
"loss": 0.17293050289154052,
"step": 2200
},
{
"epoch": 2.983761840324763,
"grad_norm": 2.674243211746216,
"learning_rate": 1.848014705211811e-05,
"loss": 0.19195995330810547,
"step": 2205
},
{
"epoch": 2.990527740189445,
"grad_norm": 1.6491018533706665,
"learning_rate": 1.837443168714203e-05,
"loss": 0.1694074034690857,
"step": 2210
},
{
"epoch": 2.997293640054127,
"grad_norm": 2.1471121311187744,
"learning_rate": 1.826884359220958e-05,
"loss": 0.1859749436378479,
"step": 2215
},
{
"epoch": 3.004059539918809,
"grad_norm": 1.5429855585098267,
"learning_rate": 1.8163384795554543e-05,
"loss": 0.15300960540771485,
"step": 2220
},
{
"epoch": 3.004059539918809,
"eval_accuracy": 0.8721999783573207,
"eval_loss": 0.30098438262939453,
"eval_runtime": 28.9741,
"eval_samples_per_second": 70.684,
"eval_steps_per_second": 4.418,
"step": 2220
},
{
"epoch": 3.010825439783491,
"grad_norm": 2.034461736679077,
"learning_rate": 1.805805732292703e-05,
"loss": 0.10880389213562011,
"step": 2225
},
{
"epoch": 3.017591339648173,
"grad_norm": 1.7300143241882324,
"learning_rate": 1.7952863197554515e-05,
"loss": 0.14824148416519164,
"step": 2230
},
{
"epoch": 3.024357239512855,
"grad_norm": 3.27923846244812,
"learning_rate": 1.7847804440103066e-05,
"loss": 0.13216495513916016,
"step": 2235
},
{
"epoch": 3.031123139377537,
"grad_norm": 2.692138671875,
"learning_rate": 1.7742883068638447e-05,
"loss": 0.11712660789489746,
"step": 2240
},
{
"epoch": 3.037889039242219,
"grad_norm": 1.943953514099121,
"learning_rate": 1.7638101098587394e-05,
"loss": 0.1616916537284851,
"step": 2245
},
{
"epoch": 3.044654939106901,
"grad_norm": 1.1808171272277832,
"learning_rate": 1.75334605426989e-05,
"loss": 0.14143142700195313,
"step": 2250
},
{
"epoch": 3.051420838971583,
"grad_norm": 1.504233956336975,
"learning_rate": 1.742896341100554e-05,
"loss": 0.16029497385025024,
"step": 2255
},
{
"epoch": 3.058186738836265,
"grad_norm": 2.27933669090271,
"learning_rate": 1.7324611710784862e-05,
"loss": 0.14582178592681885,
"step": 2260
},
{
"epoch": 3.064952638700947,
"grad_norm": 2.056269645690918,
"learning_rate": 1.7220407446520812e-05,
"loss": 0.1566435217857361,
"step": 2265
},
{
"epoch": 3.071718538565629,
"grad_norm": 2.276611566543579,
"learning_rate": 1.711635261986528e-05,
"loss": 0.19634194374084474,
"step": 2270
},
{
"epoch": 3.078484438430311,
"grad_norm": 2.3540456295013428,
"learning_rate": 1.7012449229599608e-05,
"loss": 0.13389595746994018,
"step": 2275
},
{
"epoch": 3.085250338294993,
"grad_norm": 2.257413387298584,
"learning_rate": 1.6908699271596212e-05,
"loss": 0.16535623073577882,
"step": 2280
},
{
"epoch": 3.092016238159675,
"grad_norm": 1.7060039043426514,
"learning_rate": 1.6805104738780236e-05,
"loss": 0.11289324760437011,
"step": 2285
},
{
"epoch": 3.098782138024357,
"grad_norm": 1.2384483814239502,
"learning_rate": 1.6701667621091265e-05,
"loss": 0.14795889854431152,
"step": 2290
},
{
"epoch": 3.105548037889039,
"grad_norm": 2.0937929153442383,
"learning_rate": 1.6598389905445132e-05,
"loss": 0.13685474395751954,
"step": 2295
},
{
"epoch": 3.1123139377537212,
"grad_norm": 1.1628179550170898,
"learning_rate": 1.6495273575695698e-05,
"loss": 0.10593664646148682,
"step": 2300
},
{
"epoch": 3.1190798376184032,
"grad_norm": 1.4398174285888672,
"learning_rate": 1.639232061259679e-05,
"loss": 0.1414315104484558,
"step": 2305
},
{
"epoch": 3.1258457374830853,
"grad_norm": 2.238309383392334,
"learning_rate": 1.628953299376414e-05,
"loss": 0.13256046772003174,
"step": 2310
},
{
"epoch": 3.1326116373477673,
"grad_norm": 1.6954869031906128,
"learning_rate": 1.6186912693637393e-05,
"loss": 0.13104040622711183,
"step": 2315
},
{
"epoch": 3.1393775372124493,
"grad_norm": 2.934098958969116,
"learning_rate": 1.6084461683442176e-05,
"loss": 0.13120908737182618,
"step": 2320
},
{
"epoch": 3.1461434370771313,
"grad_norm": 2.255833625793457,
"learning_rate": 1.5982181931152246e-05,
"loss": 0.12093876600265503,
"step": 2325
},
{
"epoch": 3.1529093369418133,
"grad_norm": 2.7683801651000977,
"learning_rate": 1.5880075401451678e-05,
"loss": 0.1420158863067627,
"step": 2330
},
{
"epoch": 3.1596752368064953,
"grad_norm": 1.4698132276535034,
"learning_rate": 1.5778144055697113e-05,
"loss": 0.13190693855285646,
"step": 2335
},
{
"epoch": 3.1664411366711773,
"grad_norm": 1.7112243175506592,
"learning_rate": 1.5676389851880118e-05,
"loss": 0.13382760286331177,
"step": 2340
},
{
"epoch": 3.1732070365358593,
"grad_norm": 1.1371448040008545,
"learning_rate": 1.5574814744589544e-05,
"loss": 0.15206785202026368,
"step": 2345
},
{
"epoch": 3.1799729364005414,
"grad_norm": 1.7500869035720825,
"learning_rate": 1.547342068497399e-05,
"loss": 0.11233670711517334,
"step": 2350
},
{
"epoch": 3.1867388362652234,
"grad_norm": 2.5599236488342285,
"learning_rate": 1.5372209620704324e-05,
"loss": 0.18223289251327515,
"step": 2355
},
{
"epoch": 3.1935047361299054,
"grad_norm": 2.427382707595825,
"learning_rate": 1.5271183495936273e-05,
"loss": 0.14983103275299073,
"step": 2360
},
{
"epoch": 3.2002706359945874,
"grad_norm": 2.5888164043426514,
"learning_rate": 1.5170344251273077e-05,
"loss": 0.1732112169265747,
"step": 2365
},
{
"epoch": 3.2070365358592694,
"grad_norm": 1.7968504428863525,
"learning_rate": 1.5069693823728196e-05,
"loss": 0.17494723796844483,
"step": 2370
},
{
"epoch": 3.2138024357239514,
"grad_norm": 2.2645504474639893,
"learning_rate": 1.4969234146688128e-05,
"loss": 0.15697228908538818,
"step": 2375
},
{
"epoch": 3.2205683355886334,
"grad_norm": 1.2084218263626099,
"learning_rate": 1.4868967149875257e-05,
"loss": 0.1492936134338379,
"step": 2380
},
{
"epoch": 3.2273342354533154,
"grad_norm": 1.0536212921142578,
"learning_rate": 1.4768894759310792e-05,
"loss": 0.12644997835159302,
"step": 2385
},
{
"epoch": 3.2341001353179974,
"grad_norm": 0.8510805368423462,
"learning_rate": 1.4669018897277758e-05,
"loss": 0.11946463584899902,
"step": 2390
},
{
"epoch": 3.2408660351826795,
"grad_norm": 1.5755378007888794,
"learning_rate": 1.4569341482284077e-05,
"loss": 0.12087814807891846,
"step": 2395
},
{
"epoch": 3.2476319350473615,
"grad_norm": 2.03865122795105,
"learning_rate": 1.446986442902574e-05,
"loss": 0.13381807804107665,
"step": 2400
},
{
"epoch": 3.2543978349120435,
"grad_norm": 1.0722452402114868,
"learning_rate": 1.4370589648349949e-05,
"loss": 0.16985853910446166,
"step": 2405
},
{
"epoch": 3.2611637347767255,
"grad_norm": 2.242497682571411,
"learning_rate": 1.4271519047218535e-05,
"loss": 0.11690144538879395,
"step": 2410
},
{
"epoch": 3.2679296346414075,
"grad_norm": 1.6229416131973267,
"learning_rate": 1.4172654528671203e-05,
"loss": 0.1339253306388855,
"step": 2415
},
{
"epoch": 3.2746955345060895,
"grad_norm": 2.254070997238159,
"learning_rate": 1.4073997991789078e-05,
"loss": 0.13495543003082275,
"step": 2420
},
{
"epoch": 3.2814614343707715,
"grad_norm": 2.413409948348999,
"learning_rate": 1.3975551331658144e-05,
"loss": 0.18153053522109985,
"step": 2425
},
{
"epoch": 3.2882273342354535,
"grad_norm": 1.951812505722046,
"learning_rate": 1.387731643933291e-05,
"loss": 0.17099225521087646,
"step": 2430
},
{
"epoch": 3.2949932341001356,
"grad_norm": 1.2982271909713745,
"learning_rate": 1.377929520180003e-05,
"loss": 0.13077085018157958,
"step": 2435
},
{
"epoch": 3.301759133964817,
"grad_norm": 1.6327542066574097,
"learning_rate": 1.3681489501942076e-05,
"loss": 0.17272601127624512,
"step": 2440
},
{
"epoch": 3.308525033829499,
"grad_norm": 1.3042678833007812,
"learning_rate": 1.3583901218501405e-05,
"loss": 0.14493029117584227,
"step": 2445
},
{
"epoch": 3.315290933694181,
"grad_norm": 2.0799098014831543,
"learning_rate": 1.3486532226044e-05,
"loss": 0.10982505083084107,
"step": 2450
},
{
"epoch": 3.322056833558863,
"grad_norm": 3.224494695663452,
"learning_rate": 1.3389384394923529e-05,
"loss": 0.15503368377685547,
"step": 2455
},
{
"epoch": 3.328822733423545,
"grad_norm": 1.4568700790405273,
"learning_rate": 1.3292459591245365e-05,
"loss": 0.12483901977539062,
"step": 2460
},
{
"epoch": 3.335588633288227,
"grad_norm": 1.3658103942871094,
"learning_rate": 1.3195759676830793e-05,
"loss": 0.13837902545928954,
"step": 2465
},
{
"epoch": 3.342354533152909,
"grad_norm": 1.886192798614502,
"learning_rate": 1.3099286509181185e-05,
"loss": 0.15225708484649658,
"step": 2470
},
{
"epoch": 3.349120433017591,
"grad_norm": 1.6157084703445435,
"learning_rate": 1.3003041941442373e-05,
"loss": 0.13215643167495728,
"step": 2475
},
{
"epoch": 3.3558863328822732,
"grad_norm": 2.1354048252105713,
"learning_rate": 1.2907027822369005e-05,
"loss": 0.13781520128250122,
"step": 2480
},
{
"epoch": 3.3626522327469552,
"grad_norm": 1.8297168016433716,
"learning_rate": 1.2811245996289106e-05,
"loss": 0.1387040138244629,
"step": 2485
},
{
"epoch": 3.3694181326116373,
"grad_norm": 1.8519645929336548,
"learning_rate": 1.2715698303068547e-05,
"loss": 0.1545485258102417,
"step": 2490
},
{
"epoch": 3.3761840324763193,
"grad_norm": 2.5424563884735107,
"learning_rate": 1.2620386578075804e-05,
"loss": 0.1485774040222168,
"step": 2495
},
{
"epoch": 3.3829499323410013,
"grad_norm": 2.43830943107605,
"learning_rate": 1.252531265214662e-05,
"loss": 0.15211092233657836,
"step": 2500
},
{
"epoch": 3.3897158322056833,
"grad_norm": 2.3086886405944824,
"learning_rate": 1.2430478351548908e-05,
"loss": 0.10472373962402344,
"step": 2505
},
{
"epoch": 3.3964817320703653,
"grad_norm": 2.562016248703003,
"learning_rate": 1.233588549794761e-05,
"loss": 0.15325636863708497,
"step": 2510
},
{
"epoch": 3.4032476319350473,
"grad_norm": 1.407418131828308,
"learning_rate": 1.224153590836973e-05,
"loss": 0.14291075468063355,
"step": 2515
},
{
"epoch": 3.4100135317997293,
"grad_norm": 3.0623912811279297,
"learning_rate": 1.2147431395169459e-05,
"loss": 0.15192148685455323,
"step": 2520
},
{
"epoch": 3.4167794316644113,
"grad_norm": 3.05607271194458,
"learning_rate": 1.20535737659933e-05,
"loss": 0.16197822093963624,
"step": 2525
},
{
"epoch": 3.4235453315290933,
"grad_norm": 1.8066867589950562,
"learning_rate": 1.1959964823745415e-05,
"loss": 0.13011641502380372,
"step": 2530
},
{
"epoch": 3.4303112313937754,
"grad_norm": 2.655848741531372,
"learning_rate": 1.1866606366552917e-05,
"loss": 0.11920735836029053,
"step": 2535
},
{
"epoch": 3.4370771312584574,
"grad_norm": 2.3035430908203125,
"learning_rate": 1.1773500187731426e-05,
"loss": 0.13322696685791016,
"step": 2540
},
{
"epoch": 3.4438430311231394,
"grad_norm": 3.05935001373291,
"learning_rate": 1.1680648075750486e-05,
"loss": 0.13644232749938964,
"step": 2545
},
{
"epoch": 3.4506089309878214,
"grad_norm": 1.9335527420043945,
"learning_rate": 1.1588051814199366e-05,
"loss": 0.1400376796722412,
"step": 2550
},
{
"epoch": 3.4573748308525034,
"grad_norm": 1.5316104888916016,
"learning_rate": 1.1495713181752679e-05,
"loss": 0.13692376613616944,
"step": 2555
},
{
"epoch": 3.4641407307171854,
"grad_norm": 2.612330436706543,
"learning_rate": 1.140363395213629e-05,
"loss": 0.14769126176834108,
"step": 2560
},
{
"epoch": 3.4709066305818674,
"grad_norm": 1.932735562324524,
"learning_rate": 1.1311815894093178e-05,
"loss": 0.17540912628173827,
"step": 2565
},
{
"epoch": 3.4776725304465494,
"grad_norm": 2.535245418548584,
"learning_rate": 1.1220260771349536e-05,
"loss": 0.16760412454605103,
"step": 2570
},
{
"epoch": 3.4844384303112315,
"grad_norm": 1.9871748685836792,
"learning_rate": 1.1128970342580826e-05,
"loss": 0.1639535903930664,
"step": 2575
},
{
"epoch": 3.4912043301759135,
"grad_norm": 1.5856609344482422,
"learning_rate": 1.1037946361378027e-05,
"loss": 0.13619618415832518,
"step": 2580
},
{
"epoch": 3.4979702300405955,
"grad_norm": 1.9759271144866943,
"learning_rate": 1.0947190576213961e-05,
"loss": 0.16500366926193238,
"step": 2585
},
{
"epoch": 3.5047361299052775,
"grad_norm": 1.8541984558105469,
"learning_rate": 1.085670473040967e-05,
"loss": 0.13272461891174317,
"step": 2590
},
{
"epoch": 3.5047361299052775,
"eval_accuracy": 0.8723081917541392,
"eval_loss": 0.31211531162261963,
"eval_runtime": 28.7388,
"eval_samples_per_second": 71.262,
"eval_steps_per_second": 4.454,
"step": 2590
},
{
"epoch": 3.5115020297699595,
"grad_norm": 2.263827085494995,
"learning_rate": 1.0766490562100978e-05,
"loss": 0.1429949164390564,
"step": 2595
},
{
"epoch": 3.5182679296346415,
"grad_norm": 2.104314088821411,
"learning_rate": 1.0676549804205046e-05,
"loss": 0.13890762329101564,
"step": 2600
},
{
"epoch": 3.5250338294993235,
"grad_norm": 2.041294574737549,
"learning_rate": 1.058688418438715e-05,
"loss": 0.14361535310745238,
"step": 2605
},
{
"epoch": 3.5317997293640055,
"grad_norm": 1.9828587770462036,
"learning_rate": 1.0497495425027428e-05,
"loss": 0.1366336464881897,
"step": 2610
},
{
"epoch": 3.5385656292286876,
"grad_norm": 2.9140515327453613,
"learning_rate": 1.0408385243187842e-05,
"loss": 0.13832758665084838,
"step": 2615
},
{
"epoch": 3.5453315290933696,
"grad_norm": 0.8492562174797058,
"learning_rate": 1.0319555350579168e-05,
"loss": 0.14271717071533202,
"step": 2620
},
{
"epoch": 3.5520974289580516,
"grad_norm": 1.434712529182434,
"learning_rate": 1.0231007453528149e-05,
"loss": 0.15861926078796387,
"step": 2625
},
{
"epoch": 3.558863328822733,
"grad_norm": 0.9810675382614136,
"learning_rate": 1.0142743252944666e-05,
"loss": 0.14101221561431884,
"step": 2630
},
{
"epoch": 3.565629228687415,
"grad_norm": 3.0107226371765137,
"learning_rate": 1.0054764444289134e-05,
"loss": 0.18082473278045655,
"step": 2635
},
{
"epoch": 3.572395128552097,
"grad_norm": 1.7160433530807495,
"learning_rate": 9.967072717539851e-06,
"loss": 0.15017625093460082,
"step": 2640
},
{
"epoch": 3.579161028416779,
"grad_norm": 1.3221731185913086,
"learning_rate": 9.879669757160629e-06,
"loss": 0.14145506620407106,
"step": 2645
},
{
"epoch": 3.585926928281461,
"grad_norm": 1.8209437131881714,
"learning_rate": 9.792557242068346e-06,
"loss": 0.17436301708221436,
"step": 2650
},
{
"epoch": 3.592692828146143,
"grad_norm": 2.436368227005005,
"learning_rate": 9.705736845600751e-06,
"loss": 0.15403971672058106,
"step": 2655
},
{
"epoch": 3.5994587280108252,
"grad_norm": 3.574258804321289,
"learning_rate": 9.619210235484333e-06,
"loss": 0.13450467586517334,
"step": 2660
},
{
"epoch": 3.6062246278755072,
"grad_norm": 1.341959834098816,
"learning_rate": 9.532979073802217e-06,
"loss": 0.17624325752258302,
"step": 2665
},
{
"epoch": 3.6129905277401893,
"grad_norm": 2.3629887104034424,
"learning_rate": 9.447045016962327e-06,
"loss": 0.1789612054824829,
"step": 2670
},
{
"epoch": 3.6197564276048713,
"grad_norm": 1.815994143486023,
"learning_rate": 9.361409715665487e-06,
"loss": 0.1218302607536316,
"step": 2675
},
{
"epoch": 3.6265223274695533,
"grad_norm": 2.7552428245544434,
"learning_rate": 9.276074814873778e-06,
"loss": 0.13600573539733887,
"step": 2680
},
{
"epoch": 3.6332882273342353,
"grad_norm": 1.9020435810089111,
"learning_rate": 9.191041953778868e-06,
"loss": 0.15910805463790895,
"step": 2685
},
{
"epoch": 3.6400541271989173,
"grad_norm": 2.7549989223480225,
"learning_rate": 9.106312765770616e-06,
"loss": 0.15848138332366943,
"step": 2690
},
{
"epoch": 3.6468200270635993,
"grad_norm": 2.174206256866455,
"learning_rate": 9.021888878405607e-06,
"loss": 0.16625561714172363,
"step": 2695
},
{
"epoch": 3.6535859269282813,
"grad_norm": 0.8864703178405762,
"learning_rate": 8.937771913375964e-06,
"loss": 0.11948422193527222,
"step": 2700
},
{
"epoch": 3.6603518267929633,
"grad_norm": 1.977197289466858,
"learning_rate": 8.85396348647813e-06,
"loss": 0.12896403074264526,
"step": 2705
},
{
"epoch": 3.6671177266576453,
"grad_norm": 2.512230634689331,
"learning_rate": 8.770465207581891e-06,
"loss": 0.17610430717468262,
"step": 2710
},
{
"epoch": 3.6738836265223274,
"grad_norm": 0.9213910698890686,
"learning_rate": 8.687278680599403e-06,
"loss": 0.13063844442367553,
"step": 2715
},
{
"epoch": 3.6806495263870094,
"grad_norm": 1.115181565284729,
"learning_rate": 8.604405503454399e-06,
"loss": 0.13296933174133302,
"step": 2720
},
{
"epoch": 3.6874154262516914,
"grad_norm": 1.4736729860305786,
"learning_rate": 8.521847268051525e-06,
"loss": 0.11376359462738037,
"step": 2725
},
{
"epoch": 3.6941813261163734,
"grad_norm": 2.2015531063079834,
"learning_rate": 8.439605560245703e-06,
"loss": 0.14281601905822755,
"step": 2730
},
{
"epoch": 3.7009472259810554,
"grad_norm": 2.5112862586975098,
"learning_rate": 8.357681959811733e-06,
"loss": 0.14114251136779785,
"step": 2735
},
{
"epoch": 3.7077131258457374,
"grad_norm": 0.8925549387931824,
"learning_rate": 8.276078040413878e-06,
"loss": 0.15882568359375,
"step": 2740
},
{
"epoch": 3.7144790257104194,
"grad_norm": 2.2843387126922607,
"learning_rate": 8.19479536957571e-06,
"loss": 0.11678228378295899,
"step": 2745
},
{
"epoch": 3.7212449255751014,
"grad_norm": 1.4671858549118042,
"learning_rate": 8.113835508649937e-06,
"loss": 0.14548106193542482,
"step": 2750
},
{
"epoch": 3.7280108254397835,
"grad_norm": 2.397066116333008,
"learning_rate": 8.033200012788434e-06,
"loss": 0.16912485361099244,
"step": 2755
},
{
"epoch": 3.7347767253044655,
"grad_norm": 1.9735990762710571,
"learning_rate": 7.952890430912375e-06,
"loss": 0.14341139793395996,
"step": 2760
},
{
"epoch": 3.7415426251691475,
"grad_norm": 1.5691629648208618,
"learning_rate": 7.872908305682494e-06,
"loss": 0.1423097848892212,
"step": 2765
},
{
"epoch": 3.7483085250338295,
"grad_norm": 2.6717636585235596,
"learning_rate": 7.79325517346941e-06,
"loss": 0.15379717350006103,
"step": 2770
},
{
"epoch": 3.7550744248985115,
"grad_norm": 2.239964246749878,
"learning_rate": 7.713932564324166e-06,
"loss": 0.11914598941802979,
"step": 2775
},
{
"epoch": 3.7618403247631935,
"grad_norm": 2.303854465484619,
"learning_rate": 7.634942001948783e-06,
"loss": 0.15362746715545655,
"step": 2780
},
{
"epoch": 3.7686062246278755,
"grad_norm": 0.6939072012901306,
"learning_rate": 7.55628500366706e-06,
"loss": 0.12006381750106812,
"step": 2785
},
{
"epoch": 3.7753721244925575,
"grad_norm": 1.5780203342437744,
"learning_rate": 7.477963080395328e-06,
"loss": 0.11907594203948975,
"step": 2790
},
{
"epoch": 3.7821380243572396,
"grad_norm": 2.215303421020508,
"learning_rate": 7.39997773661355e-06,
"loss": 0.13140654563903809,
"step": 2795
},
{
"epoch": 3.7889039242219216,
"grad_norm": 3.321178674697876,
"learning_rate": 7.3223304703363135e-06,
"loss": 0.12095751762390136,
"step": 2800
},
{
"epoch": 3.7956698240866036,
"grad_norm": 2.363142728805542,
"learning_rate": 7.245022773084131e-06,
"loss": 0.1488939881324768,
"step": 2805
},
{
"epoch": 3.8024357239512856,
"grad_norm": 1.52962327003479,
"learning_rate": 7.168056129854752e-06,
"loss": 0.1151964783668518,
"step": 2810
},
{
"epoch": 3.8092016238159676,
"grad_norm": 2.141237258911133,
"learning_rate": 7.091432019094627e-06,
"loss": 0.12875030040740967,
"step": 2815
},
{
"epoch": 3.8159675236806496,
"grad_norm": 2.5579235553741455,
"learning_rate": 7.0151519126705615e-06,
"loss": 0.1530660629272461,
"step": 2820
},
{
"epoch": 3.8227334235453316,
"grad_norm": 1.245128870010376,
"learning_rate": 6.939217275841353e-06,
"loss": 0.14850115776062012,
"step": 2825
},
{
"epoch": 3.8294993234100136,
"grad_norm": 2.8262155055999756,
"learning_rate": 6.863629567229757e-06,
"loss": 0.17226648330688477,
"step": 2830
},
{
"epoch": 3.8362652232746957,
"grad_norm": 2.754629373550415,
"learning_rate": 6.788390238794371e-06,
"loss": 0.1262255072593689,
"step": 2835
},
{
"epoch": 3.8430311231393777,
"grad_norm": 2.9512860774993896,
"learning_rate": 6.7135007358018115e-06,
"loss": 0.1470223307609558,
"step": 2840
},
{
"epoch": 3.8497970230040597,
"grad_norm": 2.351285696029663,
"learning_rate": 6.638962496798895e-06,
"loss": 0.1597343921661377,
"step": 2845
},
{
"epoch": 3.8565629228687417,
"grad_norm": 1.8712749481201172,
"learning_rate": 6.564776953585072e-06,
"loss": 0.10680189132690429,
"step": 2850
},
{
"epoch": 3.8633288227334237,
"grad_norm": 1.7010862827301025,
"learning_rate": 6.490945531184852e-06,
"loss": 0.18882579803466798,
"step": 2855
},
{
"epoch": 3.8700947225981057,
"grad_norm": 1.8907952308654785,
"learning_rate": 6.4174696478204815e-06,
"loss": 0.14139196872711182,
"step": 2860
},
{
"epoch": 3.8768606224627877,
"grad_norm": 2.4144303798675537,
"learning_rate": 6.344350714884673e-06,
"loss": 0.14440770149230958,
"step": 2865
},
{
"epoch": 3.8836265223274697,
"grad_norm": 1.751448392868042,
"learning_rate": 6.271590136913522e-06,
"loss": 0.15169029235839843,
"step": 2870
},
{
"epoch": 3.8903924221921518,
"grad_norm": 1.4788835048675537,
"learning_rate": 6.199189311559492e-06,
"loss": 0.13181419372558595,
"step": 2875
},
{
"epoch": 3.8971583220568338,
"grad_norm": 1.6337800025939941,
"learning_rate": 6.127149629564605e-06,
"loss": 0.13989815711975098,
"step": 2880
},
{
"epoch": 3.903924221921516,
"grad_norm": 3.5762436389923096,
"learning_rate": 6.055472474733684e-06,
"loss": 0.15823160409927367,
"step": 2885
},
{
"epoch": 3.910690121786198,
"grad_norm": 2.0639488697052,
"learning_rate": 5.984159223907823e-06,
"loss": 0.13281512260437012,
"step": 2890
},
{
"epoch": 3.91745602165088,
"grad_norm": 1.9982720613479614,
"learning_rate": 5.913211246937886e-06,
"loss": 0.13631210327148438,
"step": 2895
},
{
"epoch": 3.924221921515562,
"grad_norm": 1.6396592855453491,
"learning_rate": 5.8426299066582255e-06,
"loss": 0.16095041036605834,
"step": 2900
},
{
"epoch": 3.930987821380244,
"grad_norm": 2.556278705596924,
"learning_rate": 5.772416558860514e-06,
"loss": 0.15621118545532225,
"step": 2905
},
{
"epoch": 3.937753721244926,
"grad_norm": 1.1012766361236572,
"learning_rate": 5.702572552267655e-06,
"loss": 0.15422338247299194,
"step": 2910
},
{
"epoch": 3.944519621109608,
"grad_norm": 1.5328612327575684,
"learning_rate": 5.633099228507935e-06,
"loss": 0.1472225308418274,
"step": 2915
},
{
"epoch": 3.9512855209742894,
"grad_norm": 1.6694269180297852,
"learning_rate": 5.563997922089193e-06,
"loss": 0.10590094327926636,
"step": 2920
},
{
"epoch": 3.9580514208389714,
"grad_norm": 2.416538953781128,
"learning_rate": 5.4952699603732536e-06,
"loss": 0.15497426986694335,
"step": 2925
},
{
"epoch": 3.9648173207036534,
"grad_norm": 2.057929039001465,
"learning_rate": 5.426916663550336e-06,
"loss": 0.13179295063018798,
"step": 2930
},
{
"epoch": 3.9715832205683355,
"grad_norm": 2.1250524520874023,
"learning_rate": 5.3589393446138e-06,
"loss": 0.16140179634094237,
"step": 2935
},
{
"epoch": 3.9783491204330175,
"grad_norm": 1.9430029392242432,
"learning_rate": 5.291339309334845e-06,
"loss": 0.13807591199874877,
"step": 2940
},
{
"epoch": 3.9851150202976995,
"grad_norm": 1.6655999422073364,
"learning_rate": 5.224117856237476e-06,
"loss": 0.13647947311401368,
"step": 2945
},
{
"epoch": 3.9918809201623815,
"grad_norm": 2.6404731273651123,
"learning_rate": 5.157276276573522e-06,
"loss": 0.15894325971603393,
"step": 2950
},
{
"epoch": 3.9986468200270635,
"grad_norm": 2.222792625427246,
"learning_rate": 5.090815854297873e-06,
"loss": 0.1360340118408203,
"step": 2955
},
{
"epoch": 4.005412719891746,
"grad_norm": 3.192030191421509,
"learning_rate": 5.02473786604378e-06,
"loss": 0.12748202085494995,
"step": 2960
},
{
"epoch": 4.005412719891746,
"eval_accuracy": 0.871658911373228,
"eval_loss": 0.31526708602905273,
"eval_runtime": 28.4179,
"eval_samples_per_second": 72.067,
"eval_steps_per_second": 4.504,
"step": 2960
},
{
"epoch": 4.012178619756428,
"grad_norm": 2.5182464122772217,
"learning_rate": 4.959043581098341e-06,
"loss": 0.10190733671188354,
"step": 2965
},
{
"epoch": 4.01894451962111,
"grad_norm": 1.3963299989700317,
"learning_rate": 4.893734261378155e-06,
"loss": 0.09501264095306397,
"step": 2970
},
{
"epoch": 4.025710419485792,
"grad_norm": 2.3250949382781982,
"learning_rate": 4.828811161405017e-06,
"loss": 0.12699902057647705,
"step": 2975
},
{
"epoch": 4.032476319350474,
"grad_norm": 1.6491389274597168,
"learning_rate": 4.764275528281892e-06,
"loss": 0.15497255325317383,
"step": 2980
},
{
"epoch": 4.039242219215156,
"grad_norm": 2.221895456314087,
"learning_rate": 4.7001286016688875e-06,
"loss": 0.11853806972503662,
"step": 2985
},
{
"epoch": 4.046008119079838,
"grad_norm": 1.4312986135482788,
"learning_rate": 4.636371613759508e-06,
"loss": 0.11831581592559814,
"step": 2990
},
{
"epoch": 4.05277401894452,
"grad_norm": 1.7483818531036377,
"learning_rate": 4.573005789256937e-06,
"loss": 0.13948090076446534,
"step": 2995
},
{
"epoch": 4.059539918809202,
"grad_norm": 1.3720752000808716,
"learning_rate": 4.510032345350534e-06,
"loss": 0.14331798553466796,
"step": 3000
},
{
"epoch": 4.066305818673884,
"grad_norm": 2.186554193496704,
"learning_rate": 4.447452491692438e-06,
"loss": 0.0958407163619995,
"step": 3005
},
{
"epoch": 4.073071718538565,
"grad_norm": 1.9457632303237915,
"learning_rate": 4.385267430374373e-06,
"loss": 0.1232107400894165,
"step": 3010
},
{
"epoch": 4.079837618403247,
"grad_norm": 1.601515769958496,
"learning_rate": 4.3234783559044894e-06,
"loss": 0.15016558170318603,
"step": 3015
},
{
"epoch": 4.086603518267929,
"grad_norm": 1.8820827007293701,
"learning_rate": 4.262086455184485e-06,
"loss": 0.11488939523696899,
"step": 3020
},
{
"epoch": 4.093369418132611,
"grad_norm": 2.501232862472534,
"learning_rate": 4.201092907486759e-06,
"loss": 0.12721402645111085,
"step": 3025
},
{
"epoch": 4.100135317997293,
"grad_norm": 1.84783136844635,
"learning_rate": 4.14049888443179e-06,
"loss": 0.10910972356796264,
"step": 3030
},
{
"epoch": 4.106901217861975,
"grad_norm": 1.8950170278549194,
"learning_rate": 4.08030554996561e-06,
"loss": 0.12527880668640137,
"step": 3035
},
{
"epoch": 4.113667117726657,
"grad_norm": 2.058089256286621,
"learning_rate": 4.0205140603374465e-06,
"loss": 0.12402105331420898,
"step": 3040
},
{
"epoch": 4.120433017591339,
"grad_norm": 2.1236207485198975,
"learning_rate": 3.961125564077542e-06,
"loss": 0.138180410861969,
"step": 3045
},
{
"epoch": 4.127198917456021,
"grad_norm": 2.2066776752471924,
"learning_rate": 3.902141201975046e-06,
"loss": 0.13134593963623048,
"step": 3050
},
{
"epoch": 4.133964817320703,
"grad_norm": 2.2421181201934814,
"learning_rate": 3.843562107056145e-06,
"loss": 0.10069676637649536,
"step": 3055
},
{
"epoch": 4.140730717185385,
"grad_norm": 2.030904769897461,
"learning_rate": 3.785389404562259e-06,
"loss": 0.11818923950195312,
"step": 3060
},
{
"epoch": 4.147496617050067,
"grad_norm": 1.5498528480529785,
"learning_rate": 3.7276242119284764e-06,
"loss": 0.11917781829833984,
"step": 3065
},
{
"epoch": 4.154262516914749,
"grad_norm": 1.3159586191177368,
"learning_rate": 3.670267638762015e-06,
"loss": 0.115424644947052,
"step": 3070
},
{
"epoch": 4.161028416779431,
"grad_norm": 2.1810805797576904,
"learning_rate": 3.613320786820992e-06,
"loss": 0.1657060503959656,
"step": 3075
},
{
"epoch": 4.167794316644113,
"grad_norm": 2.0696656703948975,
"learning_rate": 3.5567847499932e-06,
"loss": 0.12218375205993652,
"step": 3080
},
{
"epoch": 4.174560216508795,
"grad_norm": 1.900846242904663,
"learning_rate": 3.5006606142751255e-06,
"loss": 0.11043152809143067,
"step": 3085
},
{
"epoch": 4.181326116373477,
"grad_norm": 0.932833194732666,
"learning_rate": 3.444949457751062e-06,
"loss": 0.11139497756958008,
"step": 3090
},
{
"epoch": 4.188092016238159,
"grad_norm": 2.4027669429779053,
"learning_rate": 3.389652350572442e-06,
"loss": 0.11497986316680908,
"step": 3095
},
{
"epoch": 4.194857916102841,
"grad_norm": 1.6453850269317627,
"learning_rate": 3.3347703549372306e-06,
"loss": 0.09540807008743286,
"step": 3100
},
{
"epoch": 4.201623815967523,
"grad_norm": 1.7752827405929565,
"learning_rate": 3.28030452506955e-06,
"loss": 0.13229609727859498,
"step": 3105
},
{
"epoch": 4.2083897158322054,
"grad_norm": 2.6079540252685547,
"learning_rate": 3.2262559071994493e-06,
"loss": 0.12988775968551636,
"step": 3110
},
{
"epoch": 4.2151556156968875,
"grad_norm": 3.8027212619781494,
"learning_rate": 3.1726255395427477e-06,
"loss": 0.12714715003967286,
"step": 3115
},
{
"epoch": 4.2219215155615695,
"grad_norm": 1.9989835023880005,
"learning_rate": 3.119414452281158e-06,
"loss": 0.13292355537414552,
"step": 3120
},
{
"epoch": 4.2286874154262515,
"grad_norm": 1.4963780641555786,
"learning_rate": 3.066623667542448e-06,
"loss": 0.12386810779571533,
"step": 3125
},
{
"epoch": 4.2354533152909335,
"grad_norm": 1.3833959102630615,
"learning_rate": 3.014254199380842e-06,
"loss": 0.11629986763000488,
"step": 3130
},
{
"epoch": 4.2422192151556155,
"grad_norm": 1.7763792276382446,
"learning_rate": 2.9623070537575097e-06,
"loss": 0.1388786554336548,
"step": 3135
},
{
"epoch": 4.2489851150202975,
"grad_norm": 3.0370209217071533,
"learning_rate": 2.910783228521269e-06,
"loss": 0.11339128017425537,
"step": 3140
},
{
"epoch": 4.2557510148849795,
"grad_norm": 2.06748628616333,
"learning_rate": 2.8596837133893993e-06,
"loss": 0.12524802684783937,
"step": 3145
},
{
"epoch": 4.2625169147496615,
"grad_norm": 3.6071295738220215,
"learning_rate": 2.8090094899286544e-06,
"loss": 0.12521322965621948,
"step": 3150
},
{
"epoch": 4.269282814614344,
"grad_norm": 2.423140525817871,
"learning_rate": 2.7587615315363748e-06,
"loss": 0.1522749900817871,
"step": 3155
},
{
"epoch": 4.276048714479026,
"grad_norm": 2.871368646621704,
"learning_rate": 2.708940803421825e-06,
"loss": 0.14141201972961426,
"step": 3160
},
{
"epoch": 4.282814614343708,
"grad_norm": 1.9419989585876465,
"learning_rate": 2.659548262587616e-06,
"loss": 0.11312873363494873,
"step": 3165
},
{
"epoch": 4.28958051420839,
"grad_norm": 2.1301467418670654,
"learning_rate": 2.6105848578113602e-06,
"loss": 0.10186258554458619,
"step": 3170
},
{
"epoch": 4.296346414073072,
"grad_norm": 2.5409653186798096,
"learning_rate": 2.5620515296274158e-06,
"loss": 0.11947979927062988,
"step": 3175
},
{
"epoch": 4.303112313937754,
"grad_norm": 2.3460474014282227,
"learning_rate": 2.513949210308833e-06,
"loss": 0.14509005546569825,
"step": 3180
},
{
"epoch": 4.309878213802436,
"grad_norm": 2.0290164947509766,
"learning_rate": 2.466278823849452e-06,
"loss": 0.14139097929000854,
"step": 3185
},
{
"epoch": 4.316644113667118,
"grad_norm": 2.5207643508911133,
"learning_rate": 2.419041285946136e-06,
"loss": 0.1314866542816162,
"step": 3190
},
{
"epoch": 4.3234100135318,
"grad_norm": 1.7020572423934937,
"learning_rate": 2.372237503981206e-06,
"loss": 0.1186953067779541,
"step": 3195
},
{
"epoch": 4.330175913396482,
"grad_norm": 1.8110159635543823,
"learning_rate": 2.325868377004986e-06,
"loss": 0.10224813222885132,
"step": 3200
},
{
"epoch": 4.336941813261164,
"grad_norm": 2.036885976791382,
"learning_rate": 2.2799347957185636e-06,
"loss": 0.11213375329971313,
"step": 3205
},
{
"epoch": 4.343707713125846,
"grad_norm": 4.074194431304932,
"learning_rate": 2.234437642456627e-06,
"loss": 0.11792477369308471,
"step": 3210
},
{
"epoch": 4.350473612990528,
"grad_norm": 1.6938408613204956,
"learning_rate": 2.1893777911705884e-06,
"loss": 0.1293353796005249,
"step": 3215
},
{
"epoch": 4.35723951285521,
"grad_norm": 2.163630723953247,
"learning_rate": 2.144756107411733e-06,
"loss": 0.09900190830230712,
"step": 3220
},
{
"epoch": 4.364005412719892,
"grad_norm": 2.110628604888916,
"learning_rate": 2.100573448314644e-06,
"loss": 0.11652176380157471,
"step": 3225
},
{
"epoch": 4.370771312584574,
"grad_norm": 1.8196337223052979,
"learning_rate": 2.0568306625806834e-06,
"loss": 0.12759778499603272,
"step": 3230
},
{
"epoch": 4.377537212449256,
"grad_norm": 2.491985321044922,
"learning_rate": 2.013528590461747e-06,
"loss": 0.14651041030883788,
"step": 3235
},
{
"epoch": 4.384303112313938,
"grad_norm": 1.4137874841690063,
"learning_rate": 1.9706680637440743e-06,
"loss": 0.10919778347015381,
"step": 3240
},
{
"epoch": 4.39106901217862,
"grad_norm": 1.2214075326919556,
"learning_rate": 1.928249905732299e-06,
"loss": 0.14215095043182374,
"step": 3245
},
{
"epoch": 4.397834912043302,
"grad_norm": 1.4114766120910645,
"learning_rate": 1.8862749312336393e-06,
"loss": 0.10450930595397949,
"step": 3250
},
{
"epoch": 4.404600811907984,
"grad_norm": 2.5714242458343506,
"learning_rate": 1.8447439465422157e-06,
"loss": 0.13852944374084472,
"step": 3255
},
{
"epoch": 4.411366711772666,
"grad_norm": 2.0099964141845703,
"learning_rate": 1.8036577494235984e-06,
"loss": 0.11386070251464844,
"step": 3260
},
{
"epoch": 4.418132611637348,
"grad_norm": 1.2030960321426392,
"learning_rate": 1.7630171290994506e-06,
"loss": 0.1246217131614685,
"step": 3265
},
{
"epoch": 4.42489851150203,
"grad_norm": 2.2948882579803467,
"learning_rate": 1.7228228662324015e-06,
"loss": 0.13796405792236327,
"step": 3270
},
{
"epoch": 4.431664411366712,
"grad_norm": 2.5403878688812256,
"learning_rate": 1.6830757329110213e-06,
"loss": 0.12515642642974853,
"step": 3275
},
{
"epoch": 4.438430311231394,
"grad_norm": 1.6545106172561646,
"learning_rate": 1.6437764926350074e-06,
"loss": 0.14116318225860597,
"step": 3280
},
{
"epoch": 4.445196211096076,
"grad_norm": 1.9403400421142578,
"learning_rate": 1.604925900300505e-06,
"loss": 0.11578557491302491,
"step": 3285
},
{
"epoch": 4.451962110960758,
"grad_norm": 1.37274968624115,
"learning_rate": 1.566524702185629e-06,
"loss": 0.12258722782135009,
"step": 3290
},
{
"epoch": 4.45872801082544,
"grad_norm": 1.0602779388427734,
"learning_rate": 1.5285736359361036e-06,
"loss": 0.13542910814285278,
"step": 3295
},
{
"epoch": 4.465493910690122,
"grad_norm": 1.2757573127746582,
"learning_rate": 1.4910734305511099e-06,
"loss": 0.0967556893825531,
"step": 3300
},
{
"epoch": 4.472259810554804,
"grad_norm": 1.6888452768325806,
"learning_rate": 1.4540248063692697e-06,
"loss": 0.14863827228546142,
"step": 3305
},
{
"epoch": 4.479025710419486,
"grad_norm": 2.008897066116333,
"learning_rate": 1.4174284750548272e-06,
"loss": 0.13936095237731932,
"step": 3310
},
{
"epoch": 4.485791610284168,
"grad_norm": 2.154024124145508,
"learning_rate": 1.3812851395839543e-06,
"loss": 0.1334796667098999,
"step": 3315
},
{
"epoch": 4.49255751014885,
"grad_norm": 1.7286471128463745,
"learning_rate": 1.345595494231261e-06,
"loss": 0.12065168619155883,
"step": 3320
},
{
"epoch": 4.499323410013532,
"grad_norm": 1.9479559659957886,
"learning_rate": 1.3103602245564728e-06,
"loss": 0.109611976146698,
"step": 3325
},
{
"epoch": 4.506089309878214,
"grad_norm": 1.8720033168792725,
"learning_rate": 1.275580007391225e-06,
"loss": 0.11663327217102051,
"step": 3330
},
{
"epoch": 4.506089309878214,
"eval_accuracy": 0.8732821123255059,
"eval_loss": 0.3330586552619934,
"eval_runtime": 28.9902,
"eval_samples_per_second": 70.645,
"eval_steps_per_second": 4.415,
"step": 3330
},
{
"epoch": 4.512855209742896,
"grad_norm": 1.7108856439590454,
"learning_rate": 1.2412555108261032e-06,
"loss": 0.10746676921844482,
"step": 3335
},
{
"epoch": 4.519621109607578,
"grad_norm": 3.029510021209717,
"learning_rate": 1.2073873941977732e-06,
"loss": 0.13037574291229248,
"step": 3340
},
{
"epoch": 4.52638700947226,
"grad_norm": 1.0683199167251587,
"learning_rate": 1.1739763080763521e-06,
"loss": 0.12728478908538818,
"step": 3345
},
{
"epoch": 4.533152909336942,
"grad_norm": 2.0898361206054688,
"learning_rate": 1.1410228942528662e-06,
"loss": 0.1245427131652832,
"step": 3350
},
{
"epoch": 4.539918809201624,
"grad_norm": 1.4553014039993286,
"learning_rate": 1.1085277857269789e-06,
"loss": 0.10855777263641357,
"step": 3355
},
{
"epoch": 4.546684709066306,
"grad_norm": 2.2443125247955322,
"learning_rate": 1.0764916066947794e-06,
"loss": 0.119187331199646,
"step": 3360
},
{
"epoch": 4.553450608930988,
"grad_norm": 1.6019930839538574,
"learning_rate": 1.0449149725368256e-06,
"loss": 0.10033470392227173,
"step": 3365
},
{
"epoch": 4.56021650879567,
"grad_norm": 1.955103874206543,
"learning_rate": 1.013798489806303e-06,
"loss": 0.12283532619476319,
"step": 3370
},
{
"epoch": 4.566982408660352,
"grad_norm": 2.290008544921875,
"learning_rate": 9.831427562173978e-07,
"loss": 0.13425986766815184,
"step": 3375
},
{
"epoch": 4.573748308525034,
"grad_norm": 1.860361099243164,
"learning_rate": 9.529483606337902e-07,
"loss": 0.14319605827331544,
"step": 3380
},
{
"epoch": 4.580514208389716,
"grad_norm": 1.678515911102295,
"learning_rate": 9.232158830573518e-07,
"loss": 0.13483121395111083,
"step": 3385
},
{
"epoch": 4.587280108254398,
"grad_norm": 2.6902356147766113,
"learning_rate": 8.939458946170043e-07,
"loss": 0.14069955348968505,
"step": 3390
},
{
"epoch": 4.59404600811908,
"grad_norm": 2.0446724891662598,
"learning_rate": 8.651389575577679e-07,
"loss": 0.11508941650390625,
"step": 3395
},
{
"epoch": 4.600811907983762,
"grad_norm": 2.3840487003326416,
"learning_rate": 8.367956252299191e-07,
"loss": 0.11468808650970459,
"step": 3400
},
{
"epoch": 4.607577807848444,
"grad_norm": 2.340008020401001,
"learning_rate": 8.089164420784085e-07,
"loss": 0.1401150941848755,
"step": 3405
},
{
"epoch": 4.614343707713126,
"grad_norm": 2.632737159729004,
"learning_rate": 7.815019436323596e-07,
"loss": 0.14080821275711058,
"step": 3410
},
{
"epoch": 4.621109607577808,
"grad_norm": 3.1295392513275146,
"learning_rate": 7.545526564948169e-07,
"loss": 0.13032712936401367,
"step": 3415
},
{
"epoch": 4.62787550744249,
"grad_norm": 2.5540883541107178,
"learning_rate": 7.280690983326121e-07,
"loss": 0.11994967460632325,
"step": 3420
},
{
"epoch": 4.634641407307172,
"grad_norm": 1.7360866069793701,
"learning_rate": 7.020517778664193e-07,
"loss": 0.12433671951293945,
"step": 3425
},
{
"epoch": 4.641407307171854,
"grad_norm": 2.0740959644317627,
"learning_rate": 6.765011948609984e-07,
"loss": 0.10409448146820069,
"step": 3430
},
{
"epoch": 4.648173207036536,
"grad_norm": 2.0580360889434814,
"learning_rate": 6.514178401155674e-07,
"loss": 0.10708047151565551,
"step": 3435
},
{
"epoch": 4.654939106901217,
"grad_norm": 1.480660319328308,
"learning_rate": 6.268021954544096e-07,
"loss": 0.09794646501541138,
"step": 3440
},
{
"epoch": 4.661705006765899,
"grad_norm": 1.638312578201294,
"learning_rate": 6.026547337175891e-07,
"loss": 0.11303920745849609,
"step": 3445
},
{
"epoch": 4.668470906630581,
"grad_norm": 1.9828901290893555,
"learning_rate": 5.789759187518895e-07,
"loss": 0.12851787805557252,
"step": 3450
},
{
"epoch": 4.675236806495263,
"grad_norm": 1.8961654901504517,
"learning_rate": 5.557662054018781e-07,
"loss": 0.12614333629608154,
"step": 3455
},
{
"epoch": 4.682002706359945,
"grad_norm": 2.1472725868225098,
"learning_rate": 5.330260395011999e-07,
"loss": 0.11661769151687622,
"step": 3460
},
{
"epoch": 4.688768606224627,
"grad_norm": 2.1781532764434814,
"learning_rate": 5.107558578639871e-07,
"loss": 0.1379758596420288,
"step": 3465
},
{
"epoch": 4.6955345060893094,
"grad_norm": 2.27986741065979,
"learning_rate": 4.889560882764905e-07,
"loss": 0.1271294116973877,
"step": 3470
},
{
"epoch": 4.7023004059539915,
"grad_norm": 1.7906800508499146,
"learning_rate": 4.6762714948884177e-07,
"loss": 0.11627352237701416,
"step": 3475
},
{
"epoch": 4.7090663058186735,
"grad_norm": 2.385115146636963,
"learning_rate": 4.4676945120702416e-07,
"loss": 0.13160851001739501,
"step": 3480
},
{
"epoch": 4.7158322056833555,
"grad_norm": 2.2625014781951904,
"learning_rate": 4.2638339408499485e-07,
"loss": 0.13013873100280762,
"step": 3485
},
{
"epoch": 4.7225981055480375,
"grad_norm": 1.924454689025879,
"learning_rate": 4.0646936971699137e-07,
"loss": 0.11710615158081054,
"step": 3490
},
{
"epoch": 4.7293640054127195,
"grad_norm": 1.7396811246871948,
"learning_rate": 3.8702776063001545e-07,
"loss": 0.11724554300308228,
"step": 3495
},
{
"epoch": 4.7361299052774015,
"grad_norm": 1.8718693256378174,
"learning_rate": 3.680589402764695e-07,
"loss": 0.12176686525344849,
"step": 3500
},
{
"epoch": 4.7428958051420835,
"grad_norm": 2.564655065536499,
"learning_rate": 3.49563273027001e-07,
"loss": 0.1310946226119995,
"step": 3505
},
{
"epoch": 4.7496617050067655,
"grad_norm": 1.7633273601531982,
"learning_rate": 3.315411141634861e-07,
"loss": 0.1348404884338379,
"step": 3510
},
{
"epoch": 4.756427604871448,
"grad_norm": 1.510155200958252,
"learning_rate": 3.1399280987222947e-07,
"loss": 0.09467498064041138,
"step": 3515
},
{
"epoch": 4.76319350473613,
"grad_norm": 1.6500880718231201,
"learning_rate": 2.969186972372806e-07,
"loss": 0.11048015356063842,
"step": 3520
},
{
"epoch": 4.769959404600812,
"grad_norm": 1.7946373224258423,
"learning_rate": 2.8031910423398924e-07,
"loss": 0.13205502033233643,
"step": 3525
},
{
"epoch": 4.776725304465494,
"grad_norm": 3.2429816722869873,
"learning_rate": 2.6419434972268786e-07,
"loss": 0.15297365188598633,
"step": 3530
},
{
"epoch": 4.783491204330176,
"grad_norm": 1.9428452253341675,
"learning_rate": 2.4854474344257474e-07,
"loss": 0.11896407604217529,
"step": 3535
},
{
"epoch": 4.790257104194858,
"grad_norm": 1.522691011428833,
"learning_rate": 2.3337058600575724e-07,
"loss": 0.10660990476608276,
"step": 3540
},
{
"epoch": 4.79702300405954,
"grad_norm": 3.232912302017212,
"learning_rate": 2.1867216889149002e-07,
"loss": 0.14744813442230226,
"step": 3545
},
{
"epoch": 4.803788903924222,
"grad_norm": 1.7124676704406738,
"learning_rate": 2.044497744405599e-07,
"loss": 0.1076197624206543,
"step": 3550
},
{
"epoch": 4.810554803788904,
"grad_norm": 2.366919994354248,
"learning_rate": 1.9070367584987093e-07,
"loss": 0.09916577339172364,
"step": 3555
},
{
"epoch": 4.817320703653586,
"grad_norm": 2.072478771209717,
"learning_rate": 1.7743413716720393e-07,
"loss": 0.1051433801651001,
"step": 3560
},
{
"epoch": 4.824086603518268,
"grad_norm": 1.390228033065796,
"learning_rate": 1.646414132861207e-07,
"loss": 0.122206711769104,
"step": 3565
},
{
"epoch": 4.83085250338295,
"grad_norm": 1.2184898853302002,
"learning_rate": 1.523257499410985e-07,
"loss": 0.11450610160827637,
"step": 3570
},
{
"epoch": 4.837618403247632,
"grad_norm": 3.555396318435669,
"learning_rate": 1.4048738370278924e-07,
"loss": 0.1428226113319397,
"step": 3575
},
{
"epoch": 4.844384303112314,
"grad_norm": 1.4550974369049072,
"learning_rate": 1.2912654197347606e-07,
"loss": 0.11183741092681884,
"step": 3580
},
{
"epoch": 4.851150202976996,
"grad_norm": 2.8932859897613525,
"learning_rate": 1.1824344298271562e-07,
"loss": 0.13104393482208251,
"step": 3585
},
{
"epoch": 4.857916102841678,
"grad_norm": 2.255586862564087,
"learning_rate": 1.0783829578314142e-07,
"loss": 0.11317722797393799,
"step": 3590
},
{
"epoch": 4.86468200270636,
"grad_norm": 2.1318538188934326,
"learning_rate": 9.791130024643935e-08,
"loss": 0.09228658676147461,
"step": 3595
},
{
"epoch": 4.871447902571042,
"grad_norm": 2.1370322704315186,
"learning_rate": 8.846264705952289e-08,
"loss": 0.13118361234664916,
"step": 3600
},
{
"epoch": 4.878213802435724,
"grad_norm": 1.9065489768981934,
"learning_rate": 7.94925177208583e-08,
"loss": 0.13511537313461303,
"step": 3605
},
{
"epoch": 4.884979702300406,
"grad_norm": 1.6746890544891357,
"learning_rate": 7.100108453698684e-08,
"loss": 0.13486255407333375,
"step": 3610
},
{
"epoch": 4.891745602165088,
"grad_norm": 1.3436098098754883,
"learning_rate": 6.298851061921074e-08,
"loss": 0.12900066375732422,
"step": 3615
},
{
"epoch": 4.89851150202977,
"grad_norm": 3.0384507179260254,
"learning_rate": 5.545494988045963e-08,
"loss": 0.13479292392730713,
"step": 3620
},
{
"epoch": 4.905277401894452,
"grad_norm": 2.010629892349243,
"learning_rate": 4.840054703233732e-08,
"loss": 0.14362790584564208,
"step": 3625
},
{
"epoch": 4.912043301759134,
"grad_norm": 2.1608994007110596,
"learning_rate": 4.182543758233515e-08,
"loss": 0.13415913581848143,
"step": 3630
},
{
"epoch": 4.918809201623816,
"grad_norm": 2.701786518096924,
"learning_rate": 3.572974783123684e-08,
"loss": 0.11590025424957276,
"step": 3635
},
{
"epoch": 4.925575101488498,
"grad_norm": 2.460134267807007,
"learning_rate": 3.011359487068987e-08,
"loss": 0.1529006242752075,
"step": 3640
},
{
"epoch": 4.93234100135318,
"grad_norm": 1.6277536153793335,
"learning_rate": 2.4977086580954544e-08,
"loss": 0.08956323266029358,
"step": 3645
},
{
"epoch": 4.939106901217862,
"grad_norm": 2.53066349029541,
"learning_rate": 2.03203216288278e-08,
"loss": 0.12726858854293824,
"step": 3650
},
{
"epoch": 4.945872801082544,
"grad_norm": 2.1464285850524902,
"learning_rate": 1.6143389465764213e-08,
"loss": 0.11074597835540771,
"step": 3655
},
{
"epoch": 4.952638700947226,
"grad_norm": 1.501103162765503,
"learning_rate": 1.2446370326132938e-08,
"loss": 0.11081557273864746,
"step": 3660
},
{
"epoch": 4.959404600811908,
"grad_norm": 1.0743277072906494,
"learning_rate": 9.22933522570224e-09,
"loss": 0.11817346811294556,
"step": 3665
},
{
"epoch": 4.96617050067659,
"grad_norm": 1.527683973312378,
"learning_rate": 6.4923459602572735e-09,
"loss": 0.12644388675689697,
"step": 3670
},
{
"epoch": 4.972936400541272,
"grad_norm": 3.4058401584625244,
"learning_rate": 4.23545510441492e-09,
"loss": 0.13416669368743897,
"step": 3675
},
{
"epoch": 4.979702300405954,
"grad_norm": 2.3298802375793457,
"learning_rate": 2.4587060106245895e-09,
"loss": 0.12568514347076415,
"step": 3680
},
{
"epoch": 4.986468200270636,
"grad_norm": 1.2374376058578491,
"learning_rate": 1.1621328083272209e-09,
"loss": 0.11424508094787597,
"step": 3685
},
{
"epoch": 4.993234100135318,
"grad_norm": 2.5257811546325684,
"learning_rate": 3.4576040329470373e-10,
"loss": 0.12172949314117432,
"step": 3690
},
{
"epoch": 5.0,
"grad_norm": 2.3782951831817627,
"learning_rate": 9.604477171909487e-12,
"loss": 0.16408214569091797,
"step": 3695
},
{
"epoch": 5.0,
"step": 3695,
"total_flos": 1.211283924332839e+18,
"train_loss": 0.1783272426731693,
"train_runtime": 3435.9244,
"train_samples_per_second": 34.398,
"train_steps_per_second": 1.075
}
],
"logging_steps": 5,
"max_steps": 3695,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 2000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.211283924332839e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}