tt1 / trainer_state.json
Kaipengm2's picture
Upload folder using huggingface_hub
fe0f7ec verified
Raw
History Blame Contribute Delete
76.5 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2030,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0024630541871921183,
"grad_norm": 1.5480924844741821,
"learning_rate": 1.9672131147540982e-05,
"loss": 2.891536331176758,
"step": 5
},
{
"epoch": 0.0049261083743842365,
"grad_norm": 0.4681042730808258,
"learning_rate": 4.4262295081967207e-05,
"loss": 2.7997177124023436,
"step": 10
},
{
"epoch": 0.007389162561576354,
"grad_norm": 0.33674123883247375,
"learning_rate": 6.885245901639344e-05,
"loss": 2.5583560943603514,
"step": 15
},
{
"epoch": 0.009852216748768473,
"grad_norm": 0.3039105534553528,
"learning_rate": 9.344262295081966e-05,
"loss": 2.397471046447754,
"step": 20
},
{
"epoch": 0.012315270935960592,
"grad_norm": 0.22720862925052643,
"learning_rate": 0.00011803278688524588,
"loss": 2.3011783599853515,
"step": 25
},
{
"epoch": 0.014778325123152709,
"grad_norm": 0.1978134661912918,
"learning_rate": 0.00014262295081967212,
"loss": 2.173269271850586,
"step": 30
},
{
"epoch": 0.017241379310344827,
"grad_norm": 0.17100553214550018,
"learning_rate": 0.00016721311475409833,
"loss": 2.0353792190551756,
"step": 35
},
{
"epoch": 0.019704433497536946,
"grad_norm": 0.22618725895881653,
"learning_rate": 0.0001918032786885246,
"loss": 1.9775619506835938,
"step": 40
},
{
"epoch": 0.022167487684729065,
"grad_norm": 0.17380759119987488,
"learning_rate": 0.0002163934426229508,
"loss": 1.9173690795898437,
"step": 45
},
{
"epoch": 0.024630541871921183,
"grad_norm": 0.20355042815208435,
"learning_rate": 0.00024098360655737703,
"loss": 1.8515590667724608,
"step": 50
},
{
"epoch": 0.027093596059113302,
"grad_norm": 0.1582382470369339,
"learning_rate": 0.00026557377049180324,
"loss": 1.821364974975586,
"step": 55
},
{
"epoch": 0.029556650246305417,
"grad_norm": 0.19577153027057648,
"learning_rate": 0.00029016393442622945,
"loss": 1.7866361618041993,
"step": 60
},
{
"epoch": 0.03201970443349754,
"grad_norm": 0.17947204411029816,
"learning_rate": 0.0002999982816514658,
"loss": 1.7368083953857423,
"step": 65
},
{
"epoch": 0.034482758620689655,
"grad_norm": 0.18685166537761688,
"learning_rate": 0.0002999877807752191,
"loss": 1.726333999633789,
"step": 70
},
{
"epoch": 0.03694581280788178,
"grad_norm": 0.20416663587093353,
"learning_rate": 0.00029996773432829007,
"loss": 1.7515968322753905,
"step": 75
},
{
"epoch": 0.03940886699507389,
"grad_norm": 0.16128577291965485,
"learning_rate": 0.00029993814358648133,
"loss": 1.7062255859375,
"step": 80
},
{
"epoch": 0.04187192118226601,
"grad_norm": 0.20772302150726318,
"learning_rate": 0.0002998990104330163,
"loss": 1.6611082077026367,
"step": 85
},
{
"epoch": 0.04433497536945813,
"grad_norm": 0.18595512211322784,
"learning_rate": 0.0002998503373584198,
"loss": 1.6555625915527343,
"step": 90
},
{
"epoch": 0.046798029556650245,
"grad_norm": 0.15224502980709076,
"learning_rate": 0.00029979212746035933,
"loss": 1.7033863067626953,
"step": 95
},
{
"epoch": 0.04926108374384237,
"grad_norm": 0.18247374892234802,
"learning_rate": 0.00029972438444344796,
"loss": 1.6605768203735352,
"step": 100
},
{
"epoch": 0.05172413793103448,
"grad_norm": 0.18089503049850464,
"learning_rate": 0.00029964711261900863,
"loss": 1.662247657775879,
"step": 105
},
{
"epoch": 0.054187192118226604,
"grad_norm": 0.16835087537765503,
"learning_rate": 0.0002995603169047997,
"loss": 1.6097984313964844,
"step": 110
},
{
"epoch": 0.05665024630541872,
"grad_norm": 0.20104758441448212,
"learning_rate": 0.00029946400282470194,
"loss": 1.6307868957519531,
"step": 115
},
{
"epoch": 0.059113300492610835,
"grad_norm": 0.1695336252450943,
"learning_rate": 0.00029935817650836706,
"loss": 1.6093120574951172,
"step": 120
},
{
"epoch": 0.06157635467980296,
"grad_norm": 0.1759975552558899,
"learning_rate": 0.00029924284469082767,
"loss": 1.5938359260559083,
"step": 125
},
{
"epoch": 0.06403940886699508,
"grad_norm": 0.17336289584636688,
"learning_rate": 0.0002991180147120685,
"loss": 1.5705293655395507,
"step": 130
},
{
"epoch": 0.0665024630541872,
"grad_norm": 0.18207502365112305,
"learning_rate": 0.0002989836945165591,
"loss": 1.6009298324584962,
"step": 135
},
{
"epoch": 0.06896551724137931,
"grad_norm": 0.1862647831439972,
"learning_rate": 0.0002988398926527488,
"loss": 1.6111469268798828,
"step": 140
},
{
"epoch": 0.07142857142857142,
"grad_norm": 0.1556750386953354,
"learning_rate": 0.0002986866182725222,
"loss": 1.5710311889648438,
"step": 145
},
{
"epoch": 0.07389162561576355,
"grad_norm": 0.17187216877937317,
"learning_rate": 0.0002985238811306167,
"loss": 1.5673177719116211,
"step": 150
},
{
"epoch": 0.07635467980295567,
"grad_norm": 0.15755166113376617,
"learning_rate": 0.0002983516915840019,
"loss": 1.5347272872924804,
"step": 155
},
{
"epoch": 0.07881773399014778,
"grad_norm": 0.19589824974536896,
"learning_rate": 0.0002981700605912205,
"loss": 1.5430304527282714,
"step": 160
},
{
"epoch": 0.0812807881773399,
"grad_norm": 0.163785919547081,
"learning_rate": 0.00029797899971169067,
"loss": 1.5797378540039062,
"step": 165
},
{
"epoch": 0.08374384236453201,
"grad_norm": 0.16454240679740906,
"learning_rate": 0.0002977785211049705,
"loss": 1.5088143348693848,
"step": 170
},
{
"epoch": 0.08620689655172414,
"grad_norm": 0.1555330604314804,
"learning_rate": 0.000297568637529984,
"loss": 1.550344467163086,
"step": 175
},
{
"epoch": 0.08866995073891626,
"grad_norm": 0.1918725222349167,
"learning_rate": 0.00029734936234420935,
"loss": 1.5254681587219239,
"step": 180
},
{
"epoch": 0.09113300492610837,
"grad_norm": 0.180306077003479,
"learning_rate": 0.00029712070950282867,
"loss": 1.5136908531188964,
"step": 185
},
{
"epoch": 0.09359605911330049,
"grad_norm": 0.19309884309768677,
"learning_rate": 0.0002968826935578396,
"loss": 1.5137138366699219,
"step": 190
},
{
"epoch": 0.0960591133004926,
"grad_norm": 0.17438358068466187,
"learning_rate": 0.0002966353296571298,
"loss": 1.522667407989502,
"step": 195
},
{
"epoch": 0.09852216748768473,
"grad_norm": 0.1779051125049591,
"learning_rate": 0.0002963786335435122,
"loss": 1.573854351043701,
"step": 200
},
{
"epoch": 0.10098522167487685,
"grad_norm": 0.1864783614873886,
"learning_rate": 0.0002961126215537237,
"loss": 1.569448184967041,
"step": 205
},
{
"epoch": 0.10344827586206896,
"grad_norm": 0.17648528516292572,
"learning_rate": 0.0002958373106173851,
"loss": 1.5183253288269043,
"step": 210
},
{
"epoch": 0.10591133004926108,
"grad_norm": 0.1765974760055542,
"learning_rate": 0.00029555271825592363,
"loss": 1.5141011238098145,
"step": 215
},
{
"epoch": 0.10837438423645321,
"grad_norm": 0.20085644721984863,
"learning_rate": 0.0002952588625814582,
"loss": 1.4946478843688964,
"step": 220
},
{
"epoch": 0.11083743842364532,
"grad_norm": 0.18272680044174194,
"learning_rate": 0.0002949557622956462,
"loss": 1.5220054626464843,
"step": 225
},
{
"epoch": 0.11330049261083744,
"grad_norm": 0.2030794322490692,
"learning_rate": 0.0002946434366884939,
"loss": 1.5377073287963867,
"step": 230
},
{
"epoch": 0.11576354679802955,
"grad_norm": 0.17467789351940155,
"learning_rate": 0.000294321905637128,
"loss": 1.4765098571777344,
"step": 235
},
{
"epoch": 0.11822660098522167,
"grad_norm": 0.19046840071678162,
"learning_rate": 0.0002939911896045316,
"loss": 1.484934139251709,
"step": 240
},
{
"epoch": 0.1206896551724138,
"grad_norm": 0.17620006203651428,
"learning_rate": 0.00029365130963824075,
"loss": 1.4772001266479493,
"step": 245
},
{
"epoch": 0.12315270935960591,
"grad_norm": 0.18947942554950714,
"learning_rate": 0.0002933022873690059,
"loss": 1.464656162261963,
"step": 250
},
{
"epoch": 0.12561576354679804,
"grad_norm": 0.180036723613739,
"learning_rate": 0.0002929441450094148,
"loss": 1.4548049926757813,
"step": 255
},
{
"epoch": 0.12807881773399016,
"grad_norm": 0.2103915512561798,
"learning_rate": 0.000292576905352479,
"loss": 1.4982961654663085,
"step": 260
},
{
"epoch": 0.13054187192118227,
"grad_norm": 0.16482633352279663,
"learning_rate": 0.00029220059177018293,
"loss": 1.5012282371520995,
"step": 265
},
{
"epoch": 0.1330049261083744,
"grad_norm": 0.18539857864379883,
"learning_rate": 0.00029181522821199705,
"loss": 1.4581329345703125,
"step": 270
},
{
"epoch": 0.1354679802955665,
"grad_norm": 0.1624497026205063,
"learning_rate": 0.000291420839203353,
"loss": 1.4343994140625,
"step": 275
},
{
"epoch": 0.13793103448275862,
"grad_norm": 0.1708306223154068,
"learning_rate": 0.0002910174498440833,
"loss": 1.4482294082641602,
"step": 280
},
{
"epoch": 0.14039408866995073,
"grad_norm": 0.17054305970668793,
"learning_rate": 0.00029060508580682333,
"loss": 1.476183032989502,
"step": 285
},
{
"epoch": 0.14285714285714285,
"grad_norm": 0.18017707765102386,
"learning_rate": 0.0002901837733353783,
"loss": 1.4687593460083008,
"step": 290
},
{
"epoch": 0.14532019704433496,
"grad_norm": 0.1773892641067505,
"learning_rate": 0.00028975353924305214,
"loss": 1.4427257537841798,
"step": 295
},
{
"epoch": 0.1477832512315271,
"grad_norm": 0.16333124041557312,
"learning_rate": 0.00028931441091094165,
"loss": 1.4356456756591798,
"step": 300
},
{
"epoch": 0.15024630541871922,
"grad_norm": 0.1586243063211441,
"learning_rate": 0.0002888664162861937,
"loss": 1.4374471664428712,
"step": 305
},
{
"epoch": 0.15270935960591134,
"grad_norm": 0.1737382858991623,
"learning_rate": 0.0002884095838802265,
"loss": 1.4098420143127441,
"step": 310
},
{
"epoch": 0.15517241379310345,
"grad_norm": 0.1684066206216812,
"learning_rate": 0.00028794394276691535,
"loss": 1.4310904502868653,
"step": 315
},
{
"epoch": 0.15763546798029557,
"grad_norm": 0.15865062177181244,
"learning_rate": 0.000287469522580742,
"loss": 1.4216425895690918,
"step": 320
},
{
"epoch": 0.16009852216748768,
"grad_norm": 0.18184268474578857,
"learning_rate": 0.00028698635351490884,
"loss": 1.4537026405334472,
"step": 325
},
{
"epoch": 0.1625615763546798,
"grad_norm": 0.15354685485363007,
"learning_rate": 0.0002864944663194173,
"loss": 1.4345867156982421,
"step": 330
},
{
"epoch": 0.16502463054187191,
"grad_norm": 0.16845227777957916,
"learning_rate": 0.00028599389229911075,
"loss": 1.3639638900756836,
"step": 335
},
{
"epoch": 0.16748768472906403,
"grad_norm": 0.17247310280799866,
"learning_rate": 0.0002854846633116824,
"loss": 1.4732954025268554,
"step": 340
},
{
"epoch": 0.16995073891625614,
"grad_norm": 0.14998307824134827,
"learning_rate": 0.0002849668117656474,
"loss": 1.407753849029541,
"step": 345
},
{
"epoch": 0.1724137931034483,
"grad_norm": 0.19805175065994263,
"learning_rate": 0.00028444037061828097,
"loss": 1.4117528915405273,
"step": 350
},
{
"epoch": 0.1748768472906404,
"grad_norm": 0.17270320653915405,
"learning_rate": 0.00028390537337352013,
"loss": 1.4096967697143554,
"step": 355
},
{
"epoch": 0.17733990147783252,
"grad_norm": 0.1353396624326706,
"learning_rate": 0.000283361854079832,
"loss": 1.4327749252319335,
"step": 360
},
{
"epoch": 0.17980295566502463,
"grad_norm": 0.15850022435188293,
"learning_rate": 0.00028280984732804667,
"loss": 1.4133735656738282,
"step": 365
},
{
"epoch": 0.18226600985221675,
"grad_norm": 0.1735433042049408,
"learning_rate": 0.00028224938824915563,
"loss": 1.4104249954223633,
"step": 370
},
{
"epoch": 0.18472906403940886,
"grad_norm": 0.1513863205909729,
"learning_rate": 0.00028168051251207626,
"loss": 1.3858678817749024,
"step": 375
},
{
"epoch": 0.18719211822660098,
"grad_norm": 0.1901034414768219,
"learning_rate": 0.00028110325632138154,
"loss": 1.3999350547790528,
"step": 380
},
{
"epoch": 0.1896551724137931,
"grad_norm": 0.17478761076927185,
"learning_rate": 0.000280517656414996,
"loss": 1.386760139465332,
"step": 385
},
{
"epoch": 0.1921182266009852,
"grad_norm": 0.16510027647018433,
"learning_rate": 0.0002799237500618577,
"loss": 1.3578205108642578,
"step": 390
},
{
"epoch": 0.19458128078817735,
"grad_norm": 0.1654566079378128,
"learning_rate": 0.000279321575059546,
"loss": 1.364866352081299,
"step": 395
},
{
"epoch": 0.19704433497536947,
"grad_norm": 0.16387440264225006,
"learning_rate": 0.0002787111697318766,
"loss": 1.4068727493286133,
"step": 400
},
{
"epoch": 0.19950738916256158,
"grad_norm": 0.16216686367988586,
"learning_rate": 0.0002780925729264621,
"loss": 1.3499589920043946,
"step": 405
},
{
"epoch": 0.2019704433497537,
"grad_norm": 0.16188740730285645,
"learning_rate": 0.0002774658240122396,
"loss": 1.4046457290649415,
"step": 410
},
{
"epoch": 0.2044334975369458,
"grad_norm": 0.15068253874778748,
"learning_rate": 0.00027683096287696565,
"loss": 1.3755802154541015,
"step": 415
},
{
"epoch": 0.20689655172413793,
"grad_norm": 0.16495083272457123,
"learning_rate": 0.0002761880299246772,
"loss": 1.3544340133666992,
"step": 420
},
{
"epoch": 0.20935960591133004,
"grad_norm": 0.1596086472272873,
"learning_rate": 0.00027553706607312024,
"loss": 1.3770217895507812,
"step": 425
},
{
"epoch": 0.21182266009852216,
"grad_norm": 0.1834026575088501,
"learning_rate": 0.0002748781127511462,
"loss": 1.3598886489868165,
"step": 430
},
{
"epoch": 0.21428571428571427,
"grad_norm": 0.17779821157455444,
"learning_rate": 0.0002742112118960747,
"loss": 1.3913217544555665,
"step": 435
},
{
"epoch": 0.21674876847290642,
"grad_norm": 0.17408569157123566,
"learning_rate": 0.00027353640595102486,
"loss": 1.3756933212280273,
"step": 440
},
{
"epoch": 0.21921182266009853,
"grad_norm": 0.15925194323062897,
"learning_rate": 0.00027285373786221413,
"loss": 1.4161248207092285,
"step": 445
},
{
"epoch": 0.22167487684729065,
"grad_norm": 0.16089503467082977,
"learning_rate": 0.0002721632510762251,
"loss": 1.3658472061157227,
"step": 450
},
{
"epoch": 0.22413793103448276,
"grad_norm": 0.16174934804439545,
"learning_rate": 0.00027146498953724014,
"loss": 1.3832536697387696,
"step": 455
},
{
"epoch": 0.22660098522167488,
"grad_norm": 0.16484405100345612,
"learning_rate": 0.0002707589976842454,
"loss": 1.360020637512207,
"step": 460
},
{
"epoch": 0.229064039408867,
"grad_norm": 0.1621040403842926,
"learning_rate": 0.00027004532044820167,
"loss": 1.325608253479004,
"step": 465
},
{
"epoch": 0.2315270935960591,
"grad_norm": 0.1713087409734726,
"learning_rate": 0.00026932400324918584,
"loss": 1.3880969047546388,
"step": 470
},
{
"epoch": 0.23399014778325122,
"grad_norm": 0.16395215690135956,
"learning_rate": 0.00026859509199349935,
"loss": 1.372953224182129,
"step": 475
},
{
"epoch": 0.23645320197044334,
"grad_norm": 0.180447056889534,
"learning_rate": 0.0002678586330707474,
"loss": 1.3375804901123047,
"step": 480
},
{
"epoch": 0.23891625615763548,
"grad_norm": 0.18610520660877228,
"learning_rate": 0.00026711467335088605,
"loss": 1.3636056900024414,
"step": 485
},
{
"epoch": 0.2413793103448276,
"grad_norm": 0.16497883200645447,
"learning_rate": 0.00026636326018123955,
"loss": 1.3901862144470214,
"step": 490
},
{
"epoch": 0.2438423645320197,
"grad_norm": 0.1603415161371231,
"learning_rate": 0.00026560444138348696,
"loss": 1.286991786956787,
"step": 495
},
{
"epoch": 0.24630541871921183,
"grad_norm": 0.19659455120563507,
"learning_rate": 0.0002648382652506188,
"loss": 1.3196422576904296,
"step": 500
},
{
"epoch": 0.24876847290640394,
"grad_norm": 0.16974778473377228,
"learning_rate": 0.0002640647805438635,
"loss": 1.3625734329223633,
"step": 505
},
{
"epoch": 0.2512315270935961,
"grad_norm": 0.1660681515932083,
"learning_rate": 0.00026328403648958395,
"loss": 1.4072540283203125,
"step": 510
},
{
"epoch": 0.2536945812807882,
"grad_norm": 0.19933657348155975,
"learning_rate": 0.00026249608277614496,
"loss": 1.351209831237793,
"step": 515
},
{
"epoch": 0.2561576354679803,
"grad_norm": 0.15851397812366486,
"learning_rate": 0.00026170096955075063,
"loss": 1.3547744750976562,
"step": 520
},
{
"epoch": 0.25862068965517243,
"grad_norm": 0.172214075922966,
"learning_rate": 0.0002608987474162531,
"loss": 1.340158462524414,
"step": 525
},
{
"epoch": 0.26108374384236455,
"grad_norm": 0.16379788517951965,
"learning_rate": 0.00026008946742793205,
"loss": 1.3426517486572265,
"step": 530
},
{
"epoch": 0.26354679802955666,
"grad_norm": 0.18204542994499207,
"learning_rate": 0.0002592731810902453,
"loss": 1.3163593292236329,
"step": 535
},
{
"epoch": 0.2660098522167488,
"grad_norm": 0.1678624302148819,
"learning_rate": 0.000258449940353551,
"loss": 1.347631072998047,
"step": 540
},
{
"epoch": 0.2684729064039409,
"grad_norm": 0.1740512102842331,
"learning_rate": 0.00025761979761080136,
"loss": 1.340871810913086,
"step": 545
},
{
"epoch": 0.270935960591133,
"grad_norm": 0.21239012479782104,
"learning_rate": 0.0002567828056942085,
"loss": 1.3281070709228515,
"step": 550
},
{
"epoch": 0.2733990147783251,
"grad_norm": 0.17629685997962952,
"learning_rate": 0.00025593901787188166,
"loss": 1.3175962448120118,
"step": 555
},
{
"epoch": 0.27586206896551724,
"grad_norm": 0.15969009697437286,
"learning_rate": 0.0002550884878444373,
"loss": 1.3408310890197754,
"step": 560
},
{
"epoch": 0.27832512315270935,
"grad_norm": 0.1572563201189041,
"learning_rate": 0.0002542312697415816,
"loss": 1.3542437553405762,
"step": 565
},
{
"epoch": 0.28078817733990147,
"grad_norm": 0.17402587831020355,
"learning_rate": 0.0002533674181186654,
"loss": 1.3361610412597655,
"step": 570
},
{
"epoch": 0.2832512315270936,
"grad_norm": 0.1775471270084381,
"learning_rate": 0.0002524969879532119,
"loss": 1.3356088638305663,
"step": 575
},
{
"epoch": 0.2857142857142857,
"grad_norm": 0.15883006155490875,
"learning_rate": 0.0002516200346414184,
"loss": 1.3263011932373048,
"step": 580
},
{
"epoch": 0.2881773399014778,
"grad_norm": 0.16821657121181488,
"learning_rate": 0.00025073661399463003,
"loss": 1.3541557312011718,
"step": 585
},
{
"epoch": 0.29064039408866993,
"grad_norm": 0.1635442078113556,
"learning_rate": 0.0002498467822357885,
"loss": 1.3555257797241211,
"step": 590
},
{
"epoch": 0.29310344827586204,
"grad_norm": 0.16849566996097565,
"learning_rate": 0.0002489505959958532,
"loss": 1.3288955688476562,
"step": 595
},
{
"epoch": 0.2955665024630542,
"grad_norm": 0.1572248786687851,
"learning_rate": 0.0002480481123101977,
"loss": 1.3337182998657227,
"step": 600
},
{
"epoch": 0.29802955665024633,
"grad_norm": 0.18141105771064758,
"learning_rate": 0.0002471393886149796,
"loss": 1.3367878913879394,
"step": 605
},
{
"epoch": 0.30049261083743845,
"grad_norm": 0.14677132666110992,
"learning_rate": 0.0002462244827434851,
"loss": 1.3215890884399415,
"step": 610
},
{
"epoch": 0.30295566502463056,
"grad_norm": 0.16903404891490936,
"learning_rate": 0.0002453034529224486,
"loss": 1.3559120178222657,
"step": 615
},
{
"epoch": 0.3054187192118227,
"grad_norm": 0.1446458399295807,
"learning_rate": 0.00024437635776834683,
"loss": 1.3678200721740723,
"step": 620
},
{
"epoch": 0.3078817733990148,
"grad_norm": 0.18641303479671478,
"learning_rate": 0.00024344325628366829,
"loss": 1.3056090354919434,
"step": 625
},
{
"epoch": 0.3103448275862069,
"grad_norm": 0.17480474710464478,
"learning_rate": 0.0002425042078531584,
"loss": 1.2561866760253906,
"step": 630
},
{
"epoch": 0.312807881773399,
"grad_norm": 0.19119766354560852,
"learning_rate": 0.00024155927224004005,
"loss": 1.3575506210327148,
"step": 635
},
{
"epoch": 0.31527093596059114,
"grad_norm": 0.16571898758411407,
"learning_rate": 0.00024060850958221007,
"loss": 1.3346943855285645,
"step": 640
},
{
"epoch": 0.31773399014778325,
"grad_norm": 0.14906178414821625,
"learning_rate": 0.00023965198038841188,
"loss": 1.337474250793457,
"step": 645
},
{
"epoch": 0.32019704433497537,
"grad_norm": 0.22248490154743195,
"learning_rate": 0.00023868974553438476,
"loss": 1.353849220275879,
"step": 650
},
{
"epoch": 0.3226600985221675,
"grad_norm": 0.18461737036705017,
"learning_rate": 0.00023772186625898938,
"loss": 1.3371232986450194,
"step": 655
},
{
"epoch": 0.3251231527093596,
"grad_norm": 0.17200283706188202,
"learning_rate": 0.00023674840416031053,
"loss": 1.3774213790893555,
"step": 660
},
{
"epoch": 0.3275862068965517,
"grad_norm": 0.1708926409482956,
"learning_rate": 0.00023576942119173677,
"loss": 1.3651674270629883,
"step": 665
},
{
"epoch": 0.33004926108374383,
"grad_norm": 0.1792900413274765,
"learning_rate": 0.00023478497965801775,
"loss": 1.2698733329772949,
"step": 670
},
{
"epoch": 0.33251231527093594,
"grad_norm": 0.1558794230222702,
"learning_rate": 0.0002337951422112988,
"loss": 1.346111011505127,
"step": 675
},
{
"epoch": 0.33497536945812806,
"grad_norm": 0.15922044217586517,
"learning_rate": 0.00023279997184713367,
"loss": 1.258944606781006,
"step": 680
},
{
"epoch": 0.3374384236453202,
"grad_norm": 0.18531592190265656,
"learning_rate": 0.00023179953190047555,
"loss": 1.278275489807129,
"step": 685
},
{
"epoch": 0.3399014778325123,
"grad_norm": 0.16116192936897278,
"learning_rate": 0.00023079388604164576,
"loss": 1.3268680572509766,
"step": 690
},
{
"epoch": 0.34236453201970446,
"grad_norm": 0.18557749688625336,
"learning_rate": 0.00022978309827228222,
"loss": 1.2669652938842773,
"step": 695
},
{
"epoch": 0.3448275862068966,
"grad_norm": 0.16507558524608612,
"learning_rate": 0.00022876723292126596,
"loss": 1.2879119873046876,
"step": 700
},
{
"epoch": 0.3472906403940887,
"grad_norm": 0.17293740808963776,
"learning_rate": 0.00022774635464062692,
"loss": 1.2909251213073731,
"step": 705
},
{
"epoch": 0.3497536945812808,
"grad_norm": 0.204181969165802,
"learning_rate": 0.00022672052840142981,
"loss": 1.2965253829956054,
"step": 710
},
{
"epoch": 0.3522167487684729,
"grad_norm": 0.17890626192092896,
"learning_rate": 0.00022568981948963858,
"loss": 1.2668149948120118,
"step": 715
},
{
"epoch": 0.35467980295566504,
"grad_norm": 0.18675513565540314,
"learning_rate": 0.00022465429350196212,
"loss": 1.3353291511535645,
"step": 720
},
{
"epoch": 0.35714285714285715,
"grad_norm": 0.1700645089149475,
"learning_rate": 0.00022361401634167898,
"loss": 1.2778221130371095,
"step": 725
},
{
"epoch": 0.35960591133004927,
"grad_norm": 0.18511028587818146,
"learning_rate": 0.00022256905421444342,
"loss": 1.3223533630371094,
"step": 730
},
{
"epoch": 0.3620689655172414,
"grad_norm": 0.1718156486749649,
"learning_rate": 0.00022151947362407203,
"loss": 1.260094165802002,
"step": 735
},
{
"epoch": 0.3645320197044335,
"grad_norm": 0.22620682418346405,
"learning_rate": 0.0002204653413683109,
"loss": 1.316976547241211,
"step": 740
},
{
"epoch": 0.3669950738916256,
"grad_norm": 0.16796429455280304,
"learning_rate": 0.00021940672453458484,
"loss": 1.2896673202514648,
"step": 745
},
{
"epoch": 0.3694581280788177,
"grad_norm": 0.17421868443489075,
"learning_rate": 0.00021834369049572754,
"loss": 1.30230712890625,
"step": 750
},
{
"epoch": 0.37192118226600984,
"grad_norm": 0.24968627095222473,
"learning_rate": 0.0002172763069056938,
"loss": 1.3192104339599608,
"step": 755
},
{
"epoch": 0.37438423645320196,
"grad_norm": 0.1759566068649292,
"learning_rate": 0.0002162046416952542,
"loss": 1.3094951629638671,
"step": 760
},
{
"epoch": 0.3768472906403941,
"grad_norm": 0.17337092757225037,
"learning_rate": 0.00021512876306767136,
"loss": 1.323702335357666,
"step": 765
},
{
"epoch": 0.3793103448275862,
"grad_norm": 0.171219140291214,
"learning_rate": 0.00021404873949435968,
"loss": 1.3060911178588868,
"step": 770
},
{
"epoch": 0.3817733990147783,
"grad_norm": 0.192230224609375,
"learning_rate": 0.00021296463971052742,
"loss": 1.266189193725586,
"step": 775
},
{
"epoch": 0.3842364532019704,
"grad_norm": 0.16826289892196655,
"learning_rate": 0.00021187653271080247,
"loss": 1.3201053619384766,
"step": 780
},
{
"epoch": 0.3866995073891626,
"grad_norm": 0.16479448974132538,
"learning_rate": 0.00021078448774484112,
"loss": 1.2861207008361817,
"step": 785
},
{
"epoch": 0.3891625615763547,
"grad_norm": 0.1776839643716812,
"learning_rate": 0.0002096885743129211,
"loss": 1.2839866638183595,
"step": 790
},
{
"epoch": 0.3916256157635468,
"grad_norm": 0.1969405859708786,
"learning_rate": 0.00020858886216151818,
"loss": 1.3057109832763671,
"step": 795
},
{
"epoch": 0.39408866995073893,
"grad_norm": 0.21142642199993134,
"learning_rate": 0.00020748542127886754,
"loss": 1.233298397064209,
"step": 800
},
{
"epoch": 0.39655172413793105,
"grad_norm": 0.17799562215805054,
"learning_rate": 0.0002063783218905094,
"loss": 1.2337217330932617,
"step": 805
},
{
"epoch": 0.39901477832512317,
"grad_norm": 0.16612942516803741,
"learning_rate": 0.00020526763445481998,
"loss": 1.2490262985229492,
"step": 810
},
{
"epoch": 0.4014778325123153,
"grad_norm": 0.19321958720684052,
"learning_rate": 0.00020415342965852682,
"loss": 1.2663690567016601,
"step": 815
},
{
"epoch": 0.4039408866995074,
"grad_norm": 0.1592732071876526,
"learning_rate": 0.00020303577841221078,
"loss": 1.2308451652526855,
"step": 820
},
{
"epoch": 0.4064039408866995,
"grad_norm": 0.2024703472852707,
"learning_rate": 0.00020191475184579271,
"loss": 1.210181427001953,
"step": 825
},
{
"epoch": 0.4088669950738916,
"grad_norm": 0.17606008052825928,
"learning_rate": 0.00020079042130400647,
"loss": 1.227010154724121,
"step": 830
},
{
"epoch": 0.41133004926108374,
"grad_norm": 0.15403436124324799,
"learning_rate": 0.0001996628583418589,
"loss": 1.2652328491210938,
"step": 835
},
{
"epoch": 0.41379310344827586,
"grad_norm": 0.17915283143520355,
"learning_rate": 0.00019853213472007534,
"loss": 1.2799953460693358,
"step": 840
},
{
"epoch": 0.41625615763546797,
"grad_norm": 0.20302283763885498,
"learning_rate": 0.00019739832240053302,
"loss": 1.2747675895690918,
"step": 845
},
{
"epoch": 0.4187192118226601,
"grad_norm": 0.17030516266822815,
"learning_rate": 0.00019626149354168092,
"loss": 1.2208684921264648,
"step": 850
},
{
"epoch": 0.4211822660098522,
"grad_norm": 0.2083371877670288,
"learning_rate": 0.00019512172049394766,
"loss": 1.2915332794189454,
"step": 855
},
{
"epoch": 0.4236453201970443,
"grad_norm": 0.19788840413093567,
"learning_rate": 0.00019397907579513687,
"loss": 1.2370890617370605,
"step": 860
},
{
"epoch": 0.42610837438423643,
"grad_norm": 0.18979376554489136,
"learning_rate": 0.00019283363216581078,
"loss": 1.2130452156066895,
"step": 865
},
{
"epoch": 0.42857142857142855,
"grad_norm": 0.17258431017398834,
"learning_rate": 0.00019168546250466193,
"loss": 1.187690258026123,
"step": 870
},
{
"epoch": 0.43103448275862066,
"grad_norm": 0.17532944679260254,
"learning_rate": 0.0001905346398838739,
"loss": 1.2625545501708983,
"step": 875
},
{
"epoch": 0.43349753694581283,
"grad_norm": 0.17488016188144684,
"learning_rate": 0.00018938123754447093,
"loss": 1.2257478713989258,
"step": 880
},
{
"epoch": 0.43596059113300495,
"grad_norm": 0.2495342642068863,
"learning_rate": 0.00018822532889165635,
"loss": 1.228908634185791,
"step": 885
},
{
"epoch": 0.43842364532019706,
"grad_norm": 0.19081343710422516,
"learning_rate": 0.00018706698749014112,
"loss": 1.2210208892822265,
"step": 890
},
{
"epoch": 0.4408866995073892,
"grad_norm": 0.18013235926628113,
"learning_rate": 0.0001859062870594622,
"loss": 1.24863920211792,
"step": 895
},
{
"epoch": 0.4433497536945813,
"grad_norm": 0.1564527302980423,
"learning_rate": 0.00018474330146929025,
"loss": 1.256571388244629,
"step": 900
},
{
"epoch": 0.4458128078817734,
"grad_norm": 0.18817733228206635,
"learning_rate": 0.00018357810473472913,
"loss": 1.2124872207641602,
"step": 905
},
{
"epoch": 0.4482758620689655,
"grad_norm": 0.19370071589946747,
"learning_rate": 0.00018241077101160494,
"loss": 1.261166477203369,
"step": 910
},
{
"epoch": 0.45073891625615764,
"grad_norm": 0.17744703590869904,
"learning_rate": 0.00018124137459174663,
"loss": 1.2019372940063477,
"step": 915
},
{
"epoch": 0.45320197044334976,
"grad_norm": 0.2393001765012741,
"learning_rate": 0.00018006998989825813,
"loss": 1.294654083251953,
"step": 920
},
{
"epoch": 0.45566502463054187,
"grad_norm": 0.16451191902160645,
"learning_rate": 0.00017889669148078155,
"loss": 1.2822996139526368,
"step": 925
},
{
"epoch": 0.458128078817734,
"grad_norm": 0.1897415667772293,
"learning_rate": 0.00017772155401075287,
"loss": 1.2190580368041992,
"step": 930
},
{
"epoch": 0.4605911330049261,
"grad_norm": 0.2263982594013214,
"learning_rate": 0.00017654465227664974,
"loss": 1.2255767822265624,
"step": 935
},
{
"epoch": 0.4630541871921182,
"grad_norm": 0.17995800077915192,
"learning_rate": 0.00017536606117923162,
"loss": 1.2123924255371095,
"step": 940
},
{
"epoch": 0.46551724137931033,
"grad_norm": 0.17838291823863983,
"learning_rate": 0.00017418585572677296,
"loss": 1.214205551147461,
"step": 945
},
{
"epoch": 0.46798029556650245,
"grad_norm": 0.2048899084329605,
"learning_rate": 0.0001730041110302895,
"loss": 1.2207268714904784,
"step": 950
},
{
"epoch": 0.47044334975369456,
"grad_norm": 0.20647300779819489,
"learning_rate": 0.00017182090229875805,
"loss": 1.1623147964477538,
"step": 955
},
{
"epoch": 0.4729064039408867,
"grad_norm": 0.19336098432540894,
"learning_rate": 0.00017063630483433006,
"loss": 1.2836381912231445,
"step": 960
},
{
"epoch": 0.4753694581280788,
"grad_norm": 0.20212344825267792,
"learning_rate": 0.00016945039402753917,
"loss": 1.1893451690673829,
"step": 965
},
{
"epoch": 0.47783251231527096,
"grad_norm": 0.17950081825256348,
"learning_rate": 0.00016826324535250308,
"loss": 1.2032151222229004,
"step": 970
},
{
"epoch": 0.4802955665024631,
"grad_norm": 0.23798821866512299,
"learning_rate": 0.00016707493436212033,
"loss": 1.2255828857421875,
"step": 975
},
{
"epoch": 0.4827586206896552,
"grad_norm": 0.19914016127586365,
"learning_rate": 0.0001658855366832621,
"loss": 1.2366257667541505,
"step": 980
},
{
"epoch": 0.4852216748768473,
"grad_norm": 0.19374223053455353,
"learning_rate": 0.00016469512801195856,
"loss": 1.244904136657715,
"step": 985
},
{
"epoch": 0.4876847290640394,
"grad_norm": 0.18931300938129425,
"learning_rate": 0.0001635037841085822,
"loss": 1.2429594039916991,
"step": 990
},
{
"epoch": 0.49014778325123154,
"grad_norm": 0.1846410185098648,
"learning_rate": 0.0001623115807930256,
"loss": 1.1956957817077636,
"step": 995
},
{
"epoch": 0.49261083743842365,
"grad_norm": 0.1838129162788391,
"learning_rate": 0.00016111859393987636,
"loss": 1.2107402801513671,
"step": 1000
},
{
"epoch": 0.49507389162561577,
"grad_norm": 0.20344485342502594,
"learning_rate": 0.00015992489947358834,
"loss": 1.1793498039245605,
"step": 1005
},
{
"epoch": 0.4975369458128079,
"grad_norm": 0.17801515758037567,
"learning_rate": 0.00015873057336364938,
"loss": 1.2365805625915527,
"step": 1010
},
{
"epoch": 0.5,
"grad_norm": 0.16365255415439606,
"learning_rate": 0.0001575356916197467,
"loss": 1.1804268836975098,
"step": 1015
},
{
"epoch": 0.5024630541871922,
"grad_norm": 0.17999859154224396,
"learning_rate": 0.00015634033028692938,
"loss": 1.1878613471984862,
"step": 1020
},
{
"epoch": 0.5049261083743842,
"grad_norm": 0.21251381933689117,
"learning_rate": 0.00015514456544076858,
"loss": 1.1664518356323241,
"step": 1025
},
{
"epoch": 0.5073891625615764,
"grad_norm": 0.19716940820217133,
"learning_rate": 0.000153948473182516,
"loss": 1.2354619979858399,
"step": 1030
},
{
"epoch": 0.5098522167487685,
"grad_norm": 0.1784435659646988,
"learning_rate": 0.00015275212963426066,
"loss": 1.2098859786987304,
"step": 1035
},
{
"epoch": 0.5123152709359606,
"grad_norm": 0.16008180379867554,
"learning_rate": 0.0001515556109340841,
"loss": 1.217707920074463,
"step": 1040
},
{
"epoch": 0.5147783251231527,
"grad_norm": 0.2051045149564743,
"learning_rate": 0.00015035899323121517,
"loss": 1.2305737495422364,
"step": 1045
},
{
"epoch": 0.5172413793103449,
"grad_norm": 0.17866598069667816,
"learning_rate": 0.00014916235268118335,
"loss": 1.2021331787109375,
"step": 1050
},
{
"epoch": 0.5197044334975369,
"grad_norm": 0.1787394881248474,
"learning_rate": 0.0001479657654409721,
"loss": 1.1698467254638671,
"step": 1055
},
{
"epoch": 0.5221674876847291,
"grad_norm": 0.19422391057014465,
"learning_rate": 0.00014676930766417231,
"loss": 1.207657241821289,
"step": 1060
},
{
"epoch": 0.5246305418719212,
"grad_norm": 0.18970182538032532,
"learning_rate": 0.0001455730554961353,
"loss": 1.1335500717163085,
"step": 1065
},
{
"epoch": 0.5270935960591133,
"grad_norm": 0.18080517649650574,
"learning_rate": 0.00014437708506912715,
"loss": 1.2182777404785157,
"step": 1070
},
{
"epoch": 0.5295566502463054,
"grad_norm": 0.20478472113609314,
"learning_rate": 0.00014318147249748313,
"loss": 1.188079261779785,
"step": 1075
},
{
"epoch": 0.5320197044334976,
"grad_norm": 0.1680886298418045,
"learning_rate": 0.00014198629387276394,
"loss": 1.1400787353515625,
"step": 1080
},
{
"epoch": 0.5344827586206896,
"grad_norm": 0.21684645116329193,
"learning_rate": 0.00014079162525891282,
"loss": 1.1481502532958985,
"step": 1085
},
{
"epoch": 0.5369458128078818,
"grad_norm": 0.20436380803585052,
"learning_rate": 0.00013959754268741458,
"loss": 1.2227207183837892,
"step": 1090
},
{
"epoch": 0.5394088669950738,
"grad_norm": 0.1777799427509308,
"learning_rate": 0.00013840412215245713,
"loss": 1.2257449150085449,
"step": 1095
},
{
"epoch": 0.541871921182266,
"grad_norm": 0.1790774166584015,
"learning_rate": 0.00013721143960609478,
"loss": 1.1976690292358398,
"step": 1100
},
{
"epoch": 0.5443349753694581,
"grad_norm": 0.18197719752788544,
"learning_rate": 0.00013601957095341454,
"loss": 1.1187736511230468,
"step": 1105
},
{
"epoch": 0.5467980295566502,
"grad_norm": 0.19829115271568298,
"learning_rate": 0.00013482859204770535,
"loss": 1.1939109802246093,
"step": 1110
},
{
"epoch": 0.5492610837438424,
"grad_norm": 0.2121841013431549,
"learning_rate": 0.00013363857868563042,
"loss": 1.2166552543640137,
"step": 1115
},
{
"epoch": 0.5517241379310345,
"grad_norm": 0.16758501529693604,
"learning_rate": 0.00013244960660240373,
"loss": 1.1762331008911133,
"step": 1120
},
{
"epoch": 0.5541871921182266,
"grad_norm": 0.19880753755569458,
"learning_rate": 0.00013126175146696987,
"loss": 1.1729421615600586,
"step": 1125
},
{
"epoch": 0.5566502463054187,
"grad_norm": 0.17931313812732697,
"learning_rate": 0.00013007508887718806,
"loss": 1.2437493324279785,
"step": 1130
},
{
"epoch": 0.5591133004926109,
"grad_norm": 0.19107341766357422,
"learning_rate": 0.0001288896943550214,
"loss": 1.1863306045532227,
"step": 1135
},
{
"epoch": 0.5615763546798029,
"grad_norm": 0.16409538686275482,
"learning_rate": 0.00012770564334173,
"loss": 1.137466812133789,
"step": 1140
},
{
"epoch": 0.5640394088669951,
"grad_norm": 0.22183066606521606,
"learning_rate": 0.00012652301119307004,
"loss": 1.2210257530212403,
"step": 1145
},
{
"epoch": 0.5665024630541872,
"grad_norm": 0.19872421026229858,
"learning_rate": 0.00012534187317449796,
"loss": 1.1910433769226074,
"step": 1150
},
{
"epoch": 0.5689655172413793,
"grad_norm": 0.1780635416507721,
"learning_rate": 0.0001241623044563801,
"loss": 1.2076547622680665,
"step": 1155
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.2000032514333725,
"learning_rate": 0.00012298438010920906,
"loss": 1.135959529876709,
"step": 1160
},
{
"epoch": 0.5738916256157636,
"grad_norm": 0.2166551947593689,
"learning_rate": 0.0001218081750988257,
"loss": 1.2029973983764648,
"step": 1165
},
{
"epoch": 0.5763546798029556,
"grad_norm": 0.2122587114572525,
"learning_rate": 0.0001206337642816483,
"loss": 1.2430255889892579,
"step": 1170
},
{
"epoch": 0.5788177339901478,
"grad_norm": 0.21357107162475586,
"learning_rate": 0.00011946122239990868,
"loss": 1.1642244338989258,
"step": 1175
},
{
"epoch": 0.5812807881773399,
"grad_norm": 0.17689666152000427,
"learning_rate": 0.00011829062407689515,
"loss": 1.1887334823608398,
"step": 1180
},
{
"epoch": 0.583743842364532,
"grad_norm": 0.16229701042175293,
"learning_rate": 0.00011712204381220345,
"loss": 1.232945156097412,
"step": 1185
},
{
"epoch": 0.5862068965517241,
"grad_norm": 0.2227810025215149,
"learning_rate": 0.00011595555597699533,
"loss": 1.1569953918457032,
"step": 1190
},
{
"epoch": 0.5886699507389163,
"grad_norm": 0.1936493217945099,
"learning_rate": 0.00011479123480926561,
"loss": 1.2667901039123535,
"step": 1195
},
{
"epoch": 0.5911330049261084,
"grad_norm": 0.18496808409690857,
"learning_rate": 0.0001136291544091172,
"loss": 1.1293052673339843,
"step": 1200
},
{
"epoch": 0.5935960591133005,
"grad_norm": 0.24020123481750488,
"learning_rate": 0.00011246938873404555,
"loss": 1.1583559036254882,
"step": 1205
},
{
"epoch": 0.5960591133004927,
"grad_norm": 0.21134839951992035,
"learning_rate": 0.00011131201159423148,
"loss": 1.2600346565246583,
"step": 1210
},
{
"epoch": 0.5985221674876847,
"grad_norm": 0.1905520111322403,
"learning_rate": 0.00011015709664784384,
"loss": 1.161275291442871,
"step": 1215
},
{
"epoch": 0.6009852216748769,
"grad_norm": 0.181251659989357,
"learning_rate": 0.00010900471739635192,
"loss": 1.1901248931884765,
"step": 1220
},
{
"epoch": 0.603448275862069,
"grad_norm": 0.19918228685855865,
"learning_rate": 0.00010785494717984748,
"loss": 1.153557586669922,
"step": 1225
},
{
"epoch": 0.6059113300492611,
"grad_norm": 0.1919139176607132,
"learning_rate": 0.0001067078591723771,
"loss": 1.1904026031494142,
"step": 1230
},
{
"epoch": 0.6083743842364532,
"grad_norm": 0.17320995032787323,
"learning_rate": 0.00010556352637728556,
"loss": 1.1903844833374024,
"step": 1235
},
{
"epoch": 0.6108374384236454,
"grad_norm": 0.18284201622009277,
"learning_rate": 0.00010442202162256927,
"loss": 1.1334742546081542,
"step": 1240
},
{
"epoch": 0.6133004926108374,
"grad_norm": 0.22032350301742554,
"learning_rate": 0.00010328341755624173,
"loss": 1.154489231109619,
"step": 1245
},
{
"epoch": 0.6157635467980296,
"grad_norm": 0.2070603370666504,
"learning_rate": 0.00010214778664170995,
"loss": 1.197908306121826,
"step": 1250
},
{
"epoch": 0.6182266009852216,
"grad_norm": 0.20040033757686615,
"learning_rate": 0.0001010152011531625,
"loss": 1.266568946838379,
"step": 1255
},
{
"epoch": 0.6206896551724138,
"grad_norm": 0.19652965664863586,
"learning_rate": 9.988573317097017e-05,
"loss": 1.1770313262939454,
"step": 1260
},
{
"epoch": 0.6231527093596059,
"grad_norm": 0.23216716945171356,
"learning_rate": 9.875945457709818e-05,
"loss": 1.125637912750244,
"step": 1265
},
{
"epoch": 0.625615763546798,
"grad_norm": 0.1960841715335846,
"learning_rate": 9.763643705053179e-05,
"loss": 1.129206085205078,
"step": 1270
},
{
"epoch": 0.6280788177339901,
"grad_norm": 0.21618643403053284,
"learning_rate": 9.651675206271436e-05,
"loss": 1.1962885856628418,
"step": 1275
},
{
"epoch": 0.6305418719211823,
"grad_norm": 0.21786664426326752,
"learning_rate": 9.54004708729988e-05,
"loss": 1.1518583297729492,
"step": 1280
},
{
"epoch": 0.6330049261083743,
"grad_norm": 0.20164407789707184,
"learning_rate": 9.428766452411233e-05,
"loss": 1.1575911521911622,
"step": 1285
},
{
"epoch": 0.6354679802955665,
"grad_norm": 0.24914158880710602,
"learning_rate": 9.317840383763513e-05,
"loss": 1.2067806243896484,
"step": 1290
},
{
"epoch": 0.6379310344827587,
"grad_norm": 0.21478649973869324,
"learning_rate": 9.207275940949341e-05,
"loss": 1.2205307006835937,
"step": 1295
},
{
"epoch": 0.6403940886699507,
"grad_norm": 0.20728349685668945,
"learning_rate": 9.097080160546616e-05,
"loss": 1.205084991455078,
"step": 1300
},
{
"epoch": 0.6428571428571429,
"grad_norm": 0.20018401741981506,
"learning_rate": 8.987260055670712e-05,
"loss": 1.1653689384460448,
"step": 1305
},
{
"epoch": 0.645320197044335,
"grad_norm": 0.1931193470954895,
"learning_rate": 8.877822615528129e-05,
"loss": 1.2364728927612305,
"step": 1310
},
{
"epoch": 0.6477832512315271,
"grad_norm": 0.20193235576152802,
"learning_rate": 8.768774804971705e-05,
"loss": 1.110558319091797,
"step": 1315
},
{
"epoch": 0.6502463054187192,
"grad_norm": 0.18794859945774078,
"learning_rate": 8.660123564057343e-05,
"loss": 1.1554101943969726,
"step": 1320
},
{
"epoch": 0.6527093596059114,
"grad_norm": 0.22068722546100616,
"learning_rate": 8.551875807602337e-05,
"loss": 1.1667287826538086,
"step": 1325
},
{
"epoch": 0.6551724137931034,
"grad_norm": 0.22684800624847412,
"learning_rate": 8.444038424745288e-05,
"loss": 1.0904604911804199,
"step": 1330
},
{
"epoch": 0.6576354679802956,
"grad_norm": 0.21784110367298126,
"learning_rate": 8.336618278507668e-05,
"loss": 1.1539340972900392,
"step": 1335
},
{
"epoch": 0.6600985221674877,
"grad_norm": 0.17376139760017395,
"learning_rate": 8.229622205357046e-05,
"loss": 1.1448725700378417,
"step": 1340
},
{
"epoch": 0.6625615763546798,
"grad_norm": 0.19699886441230774,
"learning_rate": 8.12305701477199e-05,
"loss": 1.155392551422119,
"step": 1345
},
{
"epoch": 0.6650246305418719,
"grad_norm": 0.2094455063343048,
"learning_rate": 8.016929488808697e-05,
"loss": 1.1087217330932617,
"step": 1350
},
{
"epoch": 0.6674876847290641,
"grad_norm": 0.2035309374332428,
"learning_rate": 7.911246381669394e-05,
"loss": 1.1662757873535157,
"step": 1355
},
{
"epoch": 0.6699507389162561,
"grad_norm": 0.1875893473625183,
"learning_rate": 7.80601441927243e-05,
"loss": 1.1694585800170898,
"step": 1360
},
{
"epoch": 0.6724137931034483,
"grad_norm": 0.2120542675256729,
"learning_rate": 7.70124029882427e-05,
"loss": 1.023516845703125,
"step": 1365
},
{
"epoch": 0.6748768472906403,
"grad_norm": 0.2408386766910553,
"learning_rate": 7.596930688393262e-05,
"loss": 1.0397373199462892,
"step": 1370
},
{
"epoch": 0.6773399014778325,
"grad_norm": 0.26084211468696594,
"learning_rate": 7.493092226485249e-05,
"loss": 1.09324312210083,
"step": 1375
},
{
"epoch": 0.6798029556650246,
"grad_norm": 0.18893302977085114,
"learning_rate": 7.389731521621084e-05,
"loss": 1.007591724395752,
"step": 1380
},
{
"epoch": 0.6822660098522167,
"grad_norm": 0.17730098962783813,
"learning_rate": 7.286855151916062e-05,
"loss": 1.103098487854004,
"step": 1385
},
{
"epoch": 0.6847290640394089,
"grad_norm": 0.22513681650161743,
"learning_rate": 7.184469664661256e-05,
"loss": 1.2215126037597657,
"step": 1390
},
{
"epoch": 0.687192118226601,
"grad_norm": 0.19522467255592346,
"learning_rate": 7.08258157590684e-05,
"loss": 1.088747787475586,
"step": 1395
},
{
"epoch": 0.6896551724137931,
"grad_norm": 0.23663419485092163,
"learning_rate": 6.981197370047402e-05,
"loss": 1.138905143737793,
"step": 1400
},
{
"epoch": 0.6921182266009852,
"grad_norm": 0.19911812245845795,
"learning_rate": 6.880323499409246e-05,
"loss": 1.1865692138671875,
"step": 1405
},
{
"epoch": 0.6945812807881774,
"grad_norm": 0.18014119565486908,
"learning_rate": 6.779966383839744e-05,
"loss": 1.139663028717041,
"step": 1410
},
{
"epoch": 0.6970443349753694,
"grad_norm": 0.2275015115737915,
"learning_rate": 6.680132410298802e-05,
"loss": 1.1142169952392578,
"step": 1415
},
{
"epoch": 0.6995073891625616,
"grad_norm": 0.18236544728279114,
"learning_rate": 6.580827932452333e-05,
"loss": 1.2623588562011718,
"step": 1420
},
{
"epoch": 0.7019704433497537,
"grad_norm": 0.20694544911384583,
"learning_rate": 6.482059270267914e-05,
"loss": 1.101370143890381,
"step": 1425
},
{
"epoch": 0.7044334975369458,
"grad_norm": 0.23692399263381958,
"learning_rate": 6.383832709612594e-05,
"loss": 1.138704204559326,
"step": 1430
},
{
"epoch": 0.7068965517241379,
"grad_norm": 0.21223334968090057,
"learning_rate": 6.286154501852782e-05,
"loss": 1.1377087593078614,
"step": 1435
},
{
"epoch": 0.7093596059113301,
"grad_norm": 0.19400008022785187,
"learning_rate": 6.189030863456459e-05,
"loss": 1.0621712684631348,
"step": 1440
},
{
"epoch": 0.7118226600985221,
"grad_norm": 0.2034262716770172,
"learning_rate": 6.092467975597533e-05,
"loss": 1.1555398941040038,
"step": 1445
},
{
"epoch": 0.7142857142857143,
"grad_norm": 0.18846650421619415,
"learning_rate": 5.996471983762427e-05,
"loss": 1.0480660438537597,
"step": 1450
},
{
"epoch": 0.7167487684729064,
"grad_norm": 0.19534969329833984,
"learning_rate": 5.901048997358989e-05,
"loss": 1.090800952911377,
"step": 1455
},
{
"epoch": 0.7192118226600985,
"grad_norm": 0.22339703142642975,
"learning_rate": 5.806205089327675e-05,
"loss": 1.160497283935547,
"step": 1460
},
{
"epoch": 0.7216748768472906,
"grad_norm": 0.19735680520534515,
"learning_rate": 5.7119462957550435e-05,
"loss": 1.1384342193603516,
"step": 1465
},
{
"epoch": 0.7241379310344828,
"grad_norm": 0.20986047387123108,
"learning_rate": 5.6182786154896165e-05,
"loss": 1.106205368041992,
"step": 1470
},
{
"epoch": 0.7266009852216748,
"grad_norm": 0.23361171782016754,
"learning_rate": 5.5252080097600806e-05,
"loss": 1.1555895805358887,
"step": 1475
},
{
"epoch": 0.729064039408867,
"grad_norm": 0.1888696551322937,
"learning_rate": 5.432740401795938e-05,
"loss": 1.0414323806762695,
"step": 1480
},
{
"epoch": 0.7315270935960592,
"grad_norm": 0.18796303868293762,
"learning_rate": 5.3408816764504724e-05,
"loss": 1.1281245231628418,
"step": 1485
},
{
"epoch": 0.7339901477832512,
"grad_norm": 0.18756966292858124,
"learning_rate": 5.249637679826302e-05,
"loss": 1.1479016304016114,
"step": 1490
},
{
"epoch": 0.7364532019704434,
"grad_norm": 0.21053920686244965,
"learning_rate": 5.159014218903264e-05,
"loss": 1.1450017929077148,
"step": 1495
},
{
"epoch": 0.7389162561576355,
"grad_norm": 0.2308945506811142,
"learning_rate": 5.069017061168862e-05,
"loss": 1.1324448585510254,
"step": 1500
},
{
"epoch": 0.7413793103448276,
"grad_norm": 0.2067108154296875,
"learning_rate": 4.979651934251214e-05,
"loss": 1.1060646057128907,
"step": 1505
},
{
"epoch": 0.7438423645320197,
"grad_norm": 0.19191695749759674,
"learning_rate": 4.890924525554527e-05,
"loss": 1.112843132019043,
"step": 1510
},
{
"epoch": 0.7463054187192119,
"grad_norm": 0.2178668975830078,
"learning_rate": 4.802840481897142e-05,
"loss": 1.151930046081543,
"step": 1515
},
{
"epoch": 0.7487684729064039,
"grad_norm": 0.18507851660251617,
"learning_rate": 4.715405409152145e-05,
"loss": 1.0926935195922851,
"step": 1520
},
{
"epoch": 0.7512315270935961,
"grad_norm": 0.24189282953739166,
"learning_rate": 4.6286248718906273e-05,
"loss": 1.1523954391479492,
"step": 1525
},
{
"epoch": 0.7536945812807881,
"grad_norm": 0.24427039921283722,
"learning_rate": 4.542504393027504e-05,
"loss": 1.2308634757995605,
"step": 1530
},
{
"epoch": 0.7561576354679803,
"grad_norm": 0.18832404911518097,
"learning_rate": 4.4570494534700424e-05,
"loss": 1.1625101089477539,
"step": 1535
},
{
"epoch": 0.7586206896551724,
"grad_norm": 0.1803818643093109,
"learning_rate": 4.3722654917690506e-05,
"loss": 1.1391640663146974,
"step": 1540
},
{
"epoch": 0.7610837438423645,
"grad_norm": 0.20264579355716705,
"learning_rate": 4.288157903772741e-05,
"loss": 1.0936505317687988,
"step": 1545
},
{
"epoch": 0.7635467980295566,
"grad_norm": 0.19814512133598328,
"learning_rate": 4.2047320422833305e-05,
"loss": 1.1089436531066894,
"step": 1550
},
{
"epoch": 0.7660098522167488,
"grad_norm": 0.20688515901565552,
"learning_rate": 4.1219932167163974e-05,
"loss": 1.1058782577514648,
"step": 1555
},
{
"epoch": 0.7684729064039408,
"grad_norm": 0.206832155585289,
"learning_rate": 4.039946692762928e-05,
"loss": 1.095679473876953,
"step": 1560
},
{
"epoch": 0.770935960591133,
"grad_norm": 0.2211444079875946,
"learning_rate": 3.958597692054243e-05,
"loss": 1.1729564666748047,
"step": 1565
},
{
"epoch": 0.7733990147783252,
"grad_norm": 0.21677422523498535,
"learning_rate": 3.8779513918296715e-05,
"loss": 1.1058198928833007,
"step": 1570
},
{
"epoch": 0.7758620689655172,
"grad_norm": 0.21122042834758759,
"learning_rate": 3.798012924607041e-05,
"loss": 1.1532733917236329,
"step": 1575
},
{
"epoch": 0.7783251231527094,
"grad_norm": 0.18679606914520264,
"learning_rate": 3.71878737785604e-05,
"loss": 1.048002338409424,
"step": 1580
},
{
"epoch": 0.7807881773399015,
"grad_norm": 0.19835565984249115,
"learning_rate": 3.640279793674451e-05,
"loss": 1.068185043334961,
"step": 1585
},
{
"epoch": 0.7832512315270936,
"grad_norm": 0.2534460127353668,
"learning_rate": 3.5624951684672484e-05,
"loss": 1.1565275192260742,
"step": 1590
},
{
"epoch": 0.7857142857142857,
"grad_norm": 0.21186648309230804,
"learning_rate": 3.485438452628611e-05,
"loss": 1.137218189239502,
"step": 1595
},
{
"epoch": 0.7881773399014779,
"grad_norm": 0.22528398036956787,
"learning_rate": 3.409114550226893e-05,
"loss": 1.1359540939331054,
"step": 1600
},
{
"epoch": 0.7906403940886699,
"grad_norm": 0.23210372030735016,
"learning_rate": 3.3335283186924884e-05,
"loss": 1.0747434616088867,
"step": 1605
},
{
"epoch": 0.7931034482758621,
"grad_norm": 0.20188362896442413,
"learning_rate": 3.258684568508692e-05,
"loss": 1.1836297988891602,
"step": 1610
},
{
"epoch": 0.7955665024630542,
"grad_norm": 0.1840670257806778,
"learning_rate": 3.184588062905585e-05,
"loss": 1.0538622856140136,
"step": 1615
},
{
"epoch": 0.7980295566502463,
"grad_norm": 0.24288518726825714,
"learning_rate": 3.1112435175568544e-05,
"loss": 1.2075087547302246,
"step": 1620
},
{
"epoch": 0.8004926108374384,
"grad_norm": 0.2126651257276535,
"learning_rate": 3.0386556002796907e-05,
"loss": 1.1230849266052245,
"step": 1625
},
{
"epoch": 0.8029556650246306,
"grad_norm": 0.1979854255914688,
"learning_rate": 2.9668289307377203e-05,
"loss": 1.1403017044067383,
"step": 1630
},
{
"epoch": 0.8054187192118226,
"grad_norm": 0.19223685562610626,
"learning_rate": 2.8957680801469935e-05,
"loss": 1.174708366394043,
"step": 1635
},
{
"epoch": 0.8078817733990148,
"grad_norm": 0.25866860151290894,
"learning_rate": 2.825477570985063e-05,
"loss": 1.0547324180603028,
"step": 1640
},
{
"epoch": 0.8103448275862069,
"grad_norm": 0.23177948594093323,
"learning_rate": 2.755961876703173e-05,
"loss": 1.082004451751709,
"step": 1645
},
{
"epoch": 0.812807881773399,
"grad_norm": 0.23728984594345093,
"learning_rate": 2.6872254214415417e-05,
"loss": 1.1087711334228516,
"step": 1650
},
{
"epoch": 0.8152709359605911,
"grad_norm": 0.2136259824037552,
"learning_rate": 2.6192725797477975e-05,
"loss": 1.1464354515075683,
"step": 1655
},
{
"epoch": 0.8177339901477833,
"grad_norm": 0.25212687253952026,
"learning_rate": 2.5521076762986008e-05,
"loss": 1.1384355545043945,
"step": 1660
},
{
"epoch": 0.8201970443349754,
"grad_norm": 0.21529723703861237,
"learning_rate": 2.4857349856243798e-05,
"loss": 1.1227568626403808,
"step": 1665
},
{
"epoch": 0.8226600985221675,
"grad_norm": 0.19600816071033478,
"learning_rate": 2.4201587318373054e-05,
"loss": 1.0777658462524413,
"step": 1670
},
{
"epoch": 0.8251231527093597,
"grad_norm": 0.22197465598583221,
"learning_rate": 2.355383088362454e-05,
"loss": 1.1026652336120606,
"step": 1675
},
{
"epoch": 0.8275862068965517,
"grad_norm": 0.20348075032234192,
"learning_rate": 2.2914121776722056e-05,
"loss": 1.1394284248352051,
"step": 1680
},
{
"epoch": 0.8300492610837439,
"grad_norm": 0.18871130049228668,
"learning_rate": 2.2282500710238694e-05,
"loss": 1.1343194961547851,
"step": 1685
},
{
"epoch": 0.8325123152709359,
"grad_norm": 0.22459407150745392,
"learning_rate": 2.1659007882006002e-05,
"loss": 1.1588836669921876,
"step": 1690
},
{
"epoch": 0.8349753694581281,
"grad_norm": 0.2594763934612274,
"learning_rate": 2.1043682972555475e-05,
"loss": 1.0391352653503418,
"step": 1695
},
{
"epoch": 0.8374384236453202,
"grad_norm": 0.26021042466163635,
"learning_rate": 2.0436565142593275e-05,
"loss": 1.1801486968994142,
"step": 1700
},
{
"epoch": 0.8399014778325123,
"grad_norm": 0.2193550169467926,
"learning_rate": 1.9837693030508022e-05,
"loss": 1.1204030990600586,
"step": 1705
},
{
"epoch": 0.8423645320197044,
"grad_norm": 0.20370081067085266,
"learning_rate": 1.9247104749911684e-05,
"loss": 1.0619827270507813,
"step": 1710
},
{
"epoch": 0.8448275862068966,
"grad_norm": 0.2144986093044281,
"learning_rate": 1.866483788721393e-05,
"loss": 1.0852882385253906,
"step": 1715
},
{
"epoch": 0.8472906403940886,
"grad_norm": 0.25570428371429443,
"learning_rate": 1.8090929499230054e-05,
"loss": 1.0750219345092773,
"step": 1720
},
{
"epoch": 0.8497536945812808,
"grad_norm": 0.29077258706092834,
"learning_rate": 1.7525416110822766e-05,
"loss": 1.0454490661621094,
"step": 1725
},
{
"epoch": 0.8522167487684729,
"grad_norm": 0.22639347612857819,
"learning_rate": 1.696833371257727e-05,
"loss": 1.054357147216797,
"step": 1730
},
{
"epoch": 0.854679802955665,
"grad_norm": 0.2215021699666977,
"learning_rate": 1.641971775851109e-05,
"loss": 1.1027287483215331,
"step": 1735
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.22075891494750977,
"learning_rate": 1.5879603163817667e-05,
"loss": 1.1595843315124512,
"step": 1740
},
{
"epoch": 0.8596059113300493,
"grad_norm": 0.2183758169412613,
"learning_rate": 1.534802430264404e-05,
"loss": 1.0942227363586425,
"step": 1745
},
{
"epoch": 0.8620689655172413,
"grad_norm": 0.200675830245018,
"learning_rate": 1.4825015005903434e-05,
"loss": 1.1737523078918457,
"step": 1750
},
{
"epoch": 0.8645320197044335,
"grad_norm": 0.21491429209709167,
"learning_rate": 1.4310608559122066e-05,
"loss": 1.1178741455078125,
"step": 1755
},
{
"epoch": 0.8669950738916257,
"grad_norm": 0.24912020564079285,
"learning_rate": 1.3804837700320798e-05,
"loss": 1.193882369995117,
"step": 1760
},
{
"epoch": 0.8694581280788177,
"grad_norm": 0.20939451456069946,
"learning_rate": 1.3307734617931598e-05,
"loss": 1.0629735946655274,
"step": 1765
},
{
"epoch": 0.8719211822660099,
"grad_norm": 0.20006871223449707,
"learning_rate": 1.2819330948749085e-05,
"loss": 1.1027685165405274,
"step": 1770
},
{
"epoch": 0.874384236453202,
"grad_norm": 0.1962796151638031,
"learning_rate": 1.2339657775917017e-05,
"loss": 1.0650928497314454,
"step": 1775
},
{
"epoch": 0.8768472906403941,
"grad_norm": 0.23042163252830505,
"learning_rate": 1.1868745626949927e-05,
"loss": 1.0649005889892578,
"step": 1780
},
{
"epoch": 0.8793103448275862,
"grad_norm": 0.22135719656944275,
"learning_rate": 1.140662447179066e-05,
"loss": 1.127094554901123,
"step": 1785
},
{
"epoch": 0.8817733990147784,
"grad_norm": 0.23133963346481323,
"learning_rate": 1.0953323720902679e-05,
"loss": 1.0982646942138672,
"step": 1790
},
{
"epoch": 0.8842364532019704,
"grad_norm": 0.1957237869501114,
"learning_rate": 1.0508872223398435e-05,
"loss": 1.1134883880615234,
"step": 1795
},
{
"epoch": 0.8866995073891626,
"grad_norm": 0.1999032348394394,
"learning_rate": 1.0073298265203444e-05,
"loss": 1.1821847915649415,
"step": 1800
},
{
"epoch": 0.8891625615763546,
"grad_norm": 0.23069384694099426,
"learning_rate": 9.646629567255853e-06,
"loss": 1.102443313598633,
"step": 1805
},
{
"epoch": 0.8916256157635468,
"grad_norm": 0.20853745937347412,
"learning_rate": 9.228893283742461e-06,
"loss": 1.0506309509277343,
"step": 1810
},
{
"epoch": 0.8940886699507389,
"grad_norm": 0.24035850167274475,
"learning_rate": 8.820116000370486e-06,
"loss": 1.0259310722351074,
"step": 1815
},
{
"epoch": 0.896551724137931,
"grad_norm": 0.2290971428155899,
"learning_rate": 8.420323732675504e-06,
"loss": 1.0788564682006836,
"step": 1820
},
{
"epoch": 0.8990147783251231,
"grad_norm": 0.2284020632505417,
"learning_rate": 8.02954192436589e-06,
"loss": 1.1221274375915526,
"step": 1825
},
{
"epoch": 0.9014778325123153,
"grad_norm": 0.18592405319213867,
"learning_rate": 7.647795445703392e-06,
"loss": 1.0857770919799805,
"step": 1830
},
{
"epoch": 0.9039408866995073,
"grad_norm": 0.23051847517490387,
"learning_rate": 7.27510859192047e-06,
"loss": 1.1243446350097657,
"step": 1835
},
{
"epoch": 0.9064039408866995,
"grad_norm": 0.20710034668445587,
"learning_rate": 6.911505081673929e-06,
"loss": 1.0596209526062013,
"step": 1840
},
{
"epoch": 0.9088669950738916,
"grad_norm": 0.2000182569026947,
"learning_rate": 6.557008055535628e-06,
"loss": 1.08193941116333,
"step": 1845
},
{
"epoch": 0.9113300492610837,
"grad_norm": 0.19101764261722565,
"learning_rate": 6.2116400745195084e-06,
"loss": 1.0401313781738282,
"step": 1850
},
{
"epoch": 0.9137931034482759,
"grad_norm": 0.23483411967754364,
"learning_rate": 5.875423118645906e-06,
"loss": 1.1075688362121583,
"step": 1855
},
{
"epoch": 0.916256157635468,
"grad_norm": 0.22181463241577148,
"learning_rate": 5.5483785855427235e-06,
"loss": 1.068376350402832,
"step": 1860
},
{
"epoch": 0.9187192118226601,
"grad_norm": 0.2283806949853897,
"learning_rate": 5.2305272890835026e-06,
"loss": 1.1370874404907227,
"step": 1865
},
{
"epoch": 0.9211822660098522,
"grad_norm": 0.2887274920940399,
"learning_rate": 4.9218894580629e-06,
"loss": 1.0557125091552735,
"step": 1870
},
{
"epoch": 0.9236453201970444,
"grad_norm": 0.23480144143104553,
"learning_rate": 4.6224847349092485e-06,
"loss": 1.1373451232910157,
"step": 1875
},
{
"epoch": 0.9261083743842364,
"grad_norm": 0.25622227787971497,
"learning_rate": 4.332332174434394e-06,
"loss": 1.0996230125427247,
"step": 1880
},
{
"epoch": 0.9285714285714286,
"grad_norm": 0.26269659399986267,
"learning_rate": 4.051450242621129e-06,
"loss": 1.128731918334961,
"step": 1885
},
{
"epoch": 0.9310344827586207,
"grad_norm": 0.1990889012813568,
"learning_rate": 3.7798568154478515e-06,
"loss": 1.1304019927978515,
"step": 1890
},
{
"epoch": 0.9334975369458128,
"grad_norm": 0.23430590331554413,
"learning_rate": 3.5175691777510064e-06,
"loss": 1.019291877746582,
"step": 1895
},
{
"epoch": 0.9359605911330049,
"grad_norm": 0.2118009775876999,
"learning_rate": 3.264604022124967e-06,
"loss": 1.0944375038146972,
"step": 1900
},
{
"epoch": 0.9384236453201971,
"grad_norm": 0.2094886600971222,
"learning_rate": 3.020977447859685e-06,
"loss": 1.098842239379883,
"step": 1905
},
{
"epoch": 0.9408866995073891,
"grad_norm": 0.2236415147781372,
"learning_rate": 2.7867049599161083e-06,
"loss": 1.1441487312316894,
"step": 1910
},
{
"epoch": 0.9433497536945813,
"grad_norm": 0.2752135097980499,
"learning_rate": 2.5618014679394228e-06,
"loss": 1.1376987457275392,
"step": 1915
},
{
"epoch": 0.9458128078817734,
"grad_norm": 0.21128307282924652,
"learning_rate": 2.3462812853101255e-06,
"loss": 1.0604249954223632,
"step": 1920
},
{
"epoch": 0.9482758620689655,
"grad_norm": 0.226698100566864,
"learning_rate": 2.1401581282332047e-06,
"loss": 1.1888493537902831,
"step": 1925
},
{
"epoch": 0.9507389162561576,
"grad_norm": 0.24508661031723022,
"learning_rate": 1.943445114864972e-06,
"loss": 1.0990447998046875,
"step": 1930
},
{
"epoch": 0.9532019704433498,
"grad_norm": 0.18784770369529724,
"learning_rate": 1.7561547644784457e-06,
"loss": 1.1306154251098632,
"step": 1935
},
{
"epoch": 0.9556650246305419,
"grad_norm": 0.20968304574489594,
"learning_rate": 1.578298996666505e-06,
"loss": 1.1506499290466308,
"step": 1940
},
{
"epoch": 0.958128078817734,
"grad_norm": 0.23261845111846924,
"learning_rate": 1.4098891305831972e-06,
"loss": 1.066385841369629,
"step": 1945
},
{
"epoch": 0.9605911330049262,
"grad_norm": 0.22541315853595734,
"learning_rate": 1.2509358842234974e-06,
"loss": 1.022451400756836,
"step": 1950
},
{
"epoch": 0.9630541871921182,
"grad_norm": 0.25705304741859436,
"learning_rate": 1.1014493737411701e-06,
"loss": 1.0468063354492188,
"step": 1955
},
{
"epoch": 0.9655172413793104,
"grad_norm": 0.21765866875648499,
"learning_rate": 9.61439112804868e-07,
"loss": 1.075814437866211,
"step": 1960
},
{
"epoch": 0.9679802955665024,
"grad_norm": 0.21837280690670013,
"learning_rate": 8.309140119927504e-07,
"loss": 1.1672237396240235,
"step": 1965
},
{
"epoch": 0.9704433497536946,
"grad_norm": 0.27542543411254883,
"learning_rate": 7.098823782253349e-07,
"loss": 1.134492778778076,
"step": 1970
},
{
"epoch": 0.9729064039408867,
"grad_norm": 0.24929000437259674,
"learning_rate": 5.98351914236872e-07,
"loss": 1.022162914276123,
"step": 1975
},
{
"epoch": 0.9753694581280788,
"grad_norm": 0.2033040076494217,
"learning_rate": 4.963297180851033e-07,
"loss": 1.0013565063476562,
"step": 1980
},
{
"epoch": 0.9778325123152709,
"grad_norm": 0.23415440320968628,
"learning_rate": 4.038222826995063e-07,
"loss": 1.1370353698730469,
"step": 1985
},
{
"epoch": 0.9802955665024631,
"grad_norm": 0.1983232945203781,
"learning_rate": 3.208354954680914e-07,
"loss": 1.160797691345215,
"step": 1990
},
{
"epoch": 0.9827586206896551,
"grad_norm": 0.20900410413742065,
"learning_rate": 2.4737463786273504e-07,
"loss": 1.1065376281738282,
"step": 1995
},
{
"epoch": 0.9852216748768473,
"grad_norm": 0.26447731256484985,
"learning_rate": 1.8344438510299873e-07,
"loss": 1.1547245979309082,
"step": 2000
},
{
"epoch": 0.9876847290640394,
"grad_norm": 0.20426036417484283,
"learning_rate": 1.2904880585863363e-07,
"loss": 1.046004295349121,
"step": 2005
},
{
"epoch": 0.9901477832512315,
"grad_norm": 0.23997654020786285,
"learning_rate": 8.419136199062093e-08,
"loss": 1.1322190284729003,
"step": 2010
},
{
"epoch": 0.9926108374384236,
"grad_norm": 0.21695995330810547,
"learning_rate": 4.887490833086482e-08,
"loss": 1.143213653564453,
"step": 2015
},
{
"epoch": 0.9950738916256158,
"grad_norm": 0.26213300228118896,
"learning_rate": 2.310169250050453e-08,
"loss": 1.0899484634399415,
"step": 2020
},
{
"epoch": 0.9975369458128078,
"grad_norm": 0.25474274158477783,
"learning_rate": 6.8733547668287095e-09,
"loss": 1.1756253242492676,
"step": 2025
},
{
"epoch": 1.0,
"grad_norm": 0.21422885358333588,
"learning_rate": 1.9092793892561132e-10,
"loss": 1.197043228149414,
"step": 2030
}
],
"logging_steps": 5,
"max_steps": 2030,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.951955862809477e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}