Novi-Mini-Base-Preview-1 / trainer_state.json
GGUFGuy's picture
Move checkpoint 168000 files to repository root
90645bf verified
Raw History Blame Contribute Delete
290 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.36700153736507096,
"eval_steps": 500,
"global_step": 168000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 2.1845329605063746e-06,
"grad_norm": 2.255587100982666,
"learning_rate": 0.0,
"loss": 9.08322525024414,
"step": 1
},
{
"epoch": 0.00021845329605063747,
"grad_norm": 0.8816911578178406,
"learning_rate": 0.0001485,
"loss": 8.32293516216856,
"step": 100
},
{
"epoch": 0.00043690659210127495,
"grad_norm": 1.3192135095596313,
"learning_rate": 0.00029850000000000005,
"loss": 7.127236938476562,
"step": 200
},
{
"epoch": 0.0006553598881519125,
"grad_norm": 0.6860479116439819,
"learning_rate": 0.0004485,
"loss": 5.774407958984375,
"step": 300
},
{
"epoch": 0.0008738131842025499,
"grad_norm": 0.4212465286254883,
"learning_rate": 0.0005985000000000001,
"loss": 6.07599853515625,
"step": 400
},
{
"epoch": 0.0010922664802531874,
"grad_norm": 0.42650777101516724,
"learning_rate": 0.0007485,
"loss": 5.72325439453125,
"step": 500
},
{
"epoch": 0.001310719776303825,
"grad_norm": 0.8157845139503479,
"learning_rate": 0.0008985,
"loss": 5.374498901367187,
"step": 600
},
{
"epoch": 0.0015291730723544623,
"grad_norm": 0.4278698265552521,
"learning_rate": 0.0010485,
"loss": 5.207935791015625,
"step": 700
},
{
"epoch": 0.0017476263684050998,
"grad_norm": 0.4734596610069275,
"learning_rate": 0.0011985000000000001,
"loss": 5.040882568359375,
"step": 800
},
{
"epoch": 0.0019660796644557373,
"grad_norm": 0.9601178169250488,
"learning_rate": 0.0013485,
"loss": 4.920305786132812,
"step": 900
},
{
"epoch": 0.002184532960506375,
"grad_norm": 0.44838836789131165,
"learning_rate": 0.0014985,
"loss": 4.8699337768554685,
"step": 1000
},
{
"epoch": 0.0024029862565570124,
"grad_norm": 0.4548332691192627,
"learning_rate": 0.0016485,
"loss": 4.7284173583984375,
"step": 1100
},
{
"epoch": 0.00262143955260765,
"grad_norm": 0.33958613872528076,
"learning_rate": 0.0017985000000000002,
"loss": 4.536319580078125,
"step": 1200
},
{
"epoch": 0.002839892848658287,
"grad_norm": 0.36981943249702454,
"learning_rate": 0.0019485,
"loss": 4.227422790527344,
"step": 1300
},
{
"epoch": 0.0030583461447089245,
"grad_norm": 0.318843811750412,
"learning_rate": 0.0020985,
"loss": 4.353344116210938,
"step": 1400
},
{
"epoch": 0.003276799440759562,
"grad_norm": 0.2866564989089966,
"learning_rate": 0.0022485,
"loss": 4.256644897460937,
"step": 1500
},
{
"epoch": 0.0034952527368101996,
"grad_norm": 0.3490343689918518,
"learning_rate": 0.0023985,
"loss": 4.181780700683594,
"step": 1600
},
{
"epoch": 0.003713706032860837,
"grad_norm": 0.41186341643333435,
"learning_rate": 0.0025485,
"loss": 4.137591552734375,
"step": 1700
},
{
"epoch": 0.003932159328911475,
"grad_norm": 0.2618832290172577,
"learning_rate": 0.0026985,
"loss": 4.110564270019531,
"step": 1800
},
{
"epoch": 0.004150612624962112,
"grad_norm": 0.3582988977432251,
"learning_rate": 0.0028485,
"loss": 4.033620300292969,
"step": 1900
},
{
"epoch": 0.00436906592101275,
"grad_norm": 0.29407480359077454,
"learning_rate": 0.0029985000000000003,
"loss": 4.035133056640625,
"step": 2000
},
{
"epoch": 0.004587519217063387,
"grad_norm": 0.22795212268829346,
"learning_rate": 0.003,
"loss": 3.974623107910156,
"step": 2100
},
{
"epoch": 0.004805972513114025,
"grad_norm": 0.20751744508743286,
"learning_rate": 0.003,
"loss": 3.9740194702148437,
"step": 2200
},
{
"epoch": 0.005024425809164662,
"grad_norm": 0.22139297425746918,
"learning_rate": 0.003,
"loss": 3.9664035034179688,
"step": 2300
},
{
"epoch": 0.0052428791052153,
"grad_norm": 0.2882986068725586,
"learning_rate": 0.003,
"loss": 3.894712219238281,
"step": 2400
},
{
"epoch": 0.0054613324012659364,
"grad_norm": 0.24050910770893097,
"learning_rate": 0.003,
"loss": 3.86550048828125,
"step": 2500
},
{
"epoch": 0.005679785697316574,
"grad_norm": 0.19542890787124634,
"learning_rate": 0.003,
"loss": 3.8028628540039064,
"step": 2600
},
{
"epoch": 0.0058982389933672115,
"grad_norm": 0.22483991086483002,
"learning_rate": 0.003,
"loss": 3.827759094238281,
"step": 2700
},
{
"epoch": 0.006116692289417849,
"grad_norm": 0.20627519488334656,
"learning_rate": 0.003,
"loss": 3.8404141235351563,
"step": 2800
},
{
"epoch": 0.0063351455854684866,
"grad_norm": 0.17682938277721405,
"learning_rate": 0.003,
"loss": 3.813418273925781,
"step": 2900
},
{
"epoch": 0.006553598881519124,
"grad_norm": 0.2815733253955841,
"learning_rate": 0.003,
"loss": 3.7931011962890624,
"step": 3000
},
{
"epoch": 0.006772052177569762,
"grad_norm": 0.200336754322052,
"learning_rate": 0.003,
"loss": 3.764217529296875,
"step": 3100
},
{
"epoch": 0.006990505473620399,
"grad_norm": 0.19158774614334106,
"learning_rate": 0.003,
"loss": 3.7359378051757814,
"step": 3200
},
{
"epoch": 0.007208958769671037,
"grad_norm": 0.18101529777050018,
"learning_rate": 0.003,
"loss": 3.773013000488281,
"step": 3300
},
{
"epoch": 0.007427412065721674,
"grad_norm": 0.7771575450897217,
"learning_rate": 0.003,
"loss": 3.7371151733398436,
"step": 3400
},
{
"epoch": 0.007645865361772312,
"grad_norm": 0.20689968764781952,
"learning_rate": 0.003,
"loss": 3.725350646972656,
"step": 3500
},
{
"epoch": 0.00786431865782295,
"grad_norm": 0.15840990841388702,
"learning_rate": 0.003,
"loss": 3.6854702758789064,
"step": 3600
},
{
"epoch": 0.008082771953873587,
"grad_norm": 0.18692198395729065,
"learning_rate": 0.003,
"loss": 3.7458792114257813,
"step": 3700
},
{
"epoch": 0.008301225249924224,
"grad_norm": 0.15756943821907043,
"learning_rate": 0.003,
"loss": 3.7010504150390626,
"step": 3800
},
{
"epoch": 0.008519678545974862,
"grad_norm": 0.16101692616939545,
"learning_rate": 0.003,
"loss": 3.724727783203125,
"step": 3900
},
{
"epoch": 0.0087381318420255,
"grad_norm": 0.17350031435489655,
"learning_rate": 0.003,
"loss": 3.6388311767578125,
"step": 4000
},
{
"epoch": 0.008956585138076137,
"grad_norm": 0.307090699672699,
"learning_rate": 0.003,
"loss": 3.6625192260742185,
"step": 4100
},
{
"epoch": 0.009175038434126774,
"grad_norm": 0.17554183304309845,
"learning_rate": 0.003,
"loss": 3.6276791381835936,
"step": 4200
},
{
"epoch": 0.009393491730177412,
"grad_norm": 0.19869279861450195,
"learning_rate": 0.003,
"loss": 3.6753179931640627,
"step": 4300
},
{
"epoch": 0.00961194502622805,
"grad_norm": 0.1721080243587494,
"learning_rate": 0.003,
"loss": 3.6419528198242186,
"step": 4400
},
{
"epoch": 0.009830398322278687,
"grad_norm": 0.16175901889801025,
"learning_rate": 0.003,
"loss": 3.6424264526367187,
"step": 4500
},
{
"epoch": 0.010048851618329325,
"grad_norm": 0.16099213063716888,
"learning_rate": 0.003,
"loss": 3.591856689453125,
"step": 4600
},
{
"epoch": 0.010267304914379962,
"grad_norm": 0.16880081593990326,
"learning_rate": 0.003,
"loss": 3.6297122192382814,
"step": 4700
},
{
"epoch": 0.0104857582104306,
"grad_norm": 0.17125141620635986,
"learning_rate": 0.003,
"loss": 3.578116149902344,
"step": 4800
},
{
"epoch": 0.010704211506481237,
"grad_norm": 0.24591994285583496,
"learning_rate": 0.003,
"loss": 3.5897540283203124,
"step": 4900
},
{
"epoch": 0.010922664802531873,
"grad_norm": 0.16242574155330658,
"learning_rate": 0.003,
"loss": 3.6433404541015624,
"step": 5000
},
{
"epoch": 0.01114111809858251,
"grad_norm": 0.14434590935707092,
"learning_rate": 0.003,
"loss": 3.615052490234375,
"step": 5100
},
{
"epoch": 0.011359571394633148,
"grad_norm": 0.1432674080133438,
"learning_rate": 0.003,
"loss": 3.587109375,
"step": 5200
},
{
"epoch": 0.011578024690683785,
"grad_norm": 0.1445230096578598,
"learning_rate": 0.003,
"loss": 3.5490536499023437,
"step": 5300
},
{
"epoch": 0.011796477986734423,
"grad_norm": 0.13871097564697266,
"learning_rate": 0.003,
"loss": 3.5879721069335937,
"step": 5400
},
{
"epoch": 0.01201493128278506,
"grad_norm": 0.15358251333236694,
"learning_rate": 0.003,
"loss": 3.5590310668945313,
"step": 5500
},
{
"epoch": 0.012233384578835698,
"grad_norm": 0.2113238424062729,
"learning_rate": 0.003,
"loss": 3.5402783203125,
"step": 5600
},
{
"epoch": 0.012451837874886336,
"grad_norm": 0.1575174778699875,
"learning_rate": 0.003,
"loss": 3.5282159423828126,
"step": 5700
},
{
"epoch": 0.012670291170936973,
"grad_norm": 0.2206973135471344,
"learning_rate": 0.003,
"loss": 3.5321197509765625,
"step": 5800
},
{
"epoch": 0.01288874446698761,
"grad_norm": 0.1495489776134491,
"learning_rate": 0.003,
"loss": 3.5411871337890624,
"step": 5900
},
{
"epoch": 0.013107197763038248,
"grad_norm": 0.19098496437072754,
"learning_rate": 0.003,
"loss": 3.514726257324219,
"step": 6000
},
{
"epoch": 0.013325651059088886,
"grad_norm": 0.2959604263305664,
"learning_rate": 0.003,
"loss": 3.53537109375,
"step": 6100
},
{
"epoch": 0.013544104355139523,
"grad_norm": 0.15626825392246246,
"learning_rate": 0.003,
"loss": 3.5192501831054686,
"step": 6200
},
{
"epoch": 0.01376255765119016,
"grad_norm": 0.16896215081214905,
"learning_rate": 0.003,
"loss": 3.5495986938476562,
"step": 6300
},
{
"epoch": 0.013981010947240798,
"grad_norm": 0.14018405973911285,
"learning_rate": 0.003,
"loss": 3.5275753784179686,
"step": 6400
},
{
"epoch": 0.014199464243291436,
"grad_norm": 0.15250442922115326,
"learning_rate": 0.003,
"loss": 3.5582366943359376,
"step": 6500
},
{
"epoch": 0.014417917539342073,
"grad_norm": 0.22132086753845215,
"learning_rate": 0.003,
"loss": 3.52295654296875,
"step": 6600
},
{
"epoch": 0.01463637083539271,
"grad_norm": 0.1595185250043869,
"learning_rate": 0.003,
"loss": 3.5543255615234375,
"step": 6700
},
{
"epoch": 0.014854824131443348,
"grad_norm": 0.1608801633119583,
"learning_rate": 0.003,
"loss": 3.533863830566406,
"step": 6800
},
{
"epoch": 0.015073277427493986,
"grad_norm": 0.15735548734664917,
"learning_rate": 0.003,
"loss": 3.5444400024414064,
"step": 6900
},
{
"epoch": 0.015291730723544623,
"grad_norm": 0.15741421282291412,
"learning_rate": 0.003,
"loss": 3.5775491333007814,
"step": 7000
},
{
"epoch": 0.015510184019595261,
"grad_norm": 0.1459587663412094,
"learning_rate": 0.003,
"loss": 3.5826260375976564,
"step": 7100
},
{
"epoch": 0.0157286373156459,
"grad_norm": 0.17190411686897278,
"learning_rate": 0.003,
"loss": 3.5276693725585937,
"step": 7200
},
{
"epoch": 0.015947090611696534,
"grad_norm": 0.13370057940483093,
"learning_rate": 0.003,
"loss": 3.5145721435546875,
"step": 7300
},
{
"epoch": 0.016165543907747174,
"grad_norm": 0.1334085464477539,
"learning_rate": 0.003,
"loss": 3.528052978515625,
"step": 7400
},
{
"epoch": 0.01638399720379781,
"grad_norm": 0.1612652987241745,
"learning_rate": 0.003,
"loss": 3.4622488403320313,
"step": 7500
},
{
"epoch": 0.01660245049984845,
"grad_norm": 0.15845361351966858,
"learning_rate": 0.003,
"loss": 3.484794921875,
"step": 7600
},
{
"epoch": 0.016820903795899084,
"grad_norm": 0.14460201561450958,
"learning_rate": 0.003,
"loss": 3.488492126464844,
"step": 7700
},
{
"epoch": 0.017039357091949724,
"grad_norm": 0.0004208660975564271,
"learning_rate": 0.003,
"loss": 3.0585507202148436,
"step": 7800
},
{
"epoch": 0.01725781038800036,
"grad_norm": 0.16746053099632263,
"learning_rate": 0.003,
"loss": 3.4115805053710937,
"step": 7900
},
{
"epoch": 0.017476263684051,
"grad_norm": 0.12478704750537872,
"learning_rate": 0.003,
"loss": 3.5240304565429685,
"step": 8000
},
{
"epoch": 0.017694716980101634,
"grad_norm": 0.15716327726840973,
"learning_rate": 0.003,
"loss": 3.4820321655273436,
"step": 8100
},
{
"epoch": 0.017913170276152274,
"grad_norm": 0.13654641807079315,
"learning_rate": 0.003,
"loss": 3.515375061035156,
"step": 8200
},
{
"epoch": 0.01813162357220291,
"grad_norm": 0.13003572821617126,
"learning_rate": 0.003,
"loss": 3.493472900390625,
"step": 8300
},
{
"epoch": 0.01835007686825355,
"grad_norm": 0.14350499212741852,
"learning_rate": 0.003,
"loss": 3.4743597412109377,
"step": 8400
},
{
"epoch": 0.018568530164304185,
"grad_norm": 0.13438019156455994,
"learning_rate": 0.003,
"loss": 3.4803411865234377,
"step": 8500
},
{
"epoch": 0.018786983460354824,
"grad_norm": 0.13640356063842773,
"learning_rate": 0.003,
"loss": 3.486788635253906,
"step": 8600
},
{
"epoch": 0.01900543675640546,
"grad_norm": 0.18971936404705048,
"learning_rate": 0.003,
"loss": 3.482457580566406,
"step": 8700
},
{
"epoch": 0.0192238900524561,
"grad_norm": 0.1473148614168167,
"learning_rate": 0.003,
"loss": 3.456876220703125,
"step": 8800
},
{
"epoch": 0.019442343348506735,
"grad_norm": 0.18119043111801147,
"learning_rate": 0.003,
"loss": 3.462286376953125,
"step": 8900
},
{
"epoch": 0.019660796644557374,
"grad_norm": 0.13301891088485718,
"learning_rate": 0.003,
"loss": 3.4378567504882813,
"step": 9000
},
{
"epoch": 0.01987924994060801,
"grad_norm": 0.12677288055419922,
"learning_rate": 0.003,
"loss": 3.4668447875976565,
"step": 9100
},
{
"epoch": 0.02009770323665865,
"grad_norm": 0.1373668909072876,
"learning_rate": 0.003,
"loss": 3.443948974609375,
"step": 9200
},
{
"epoch": 0.020316156532709285,
"grad_norm": 0.14534036815166473,
"learning_rate": 0.003,
"loss": 3.380128173828125,
"step": 9300
},
{
"epoch": 0.020534609828759924,
"grad_norm": 0.1568584442138672,
"learning_rate": 0.003,
"loss": 3.44497314453125,
"step": 9400
},
{
"epoch": 0.02075306312481056,
"grad_norm": 0.1420242041349411,
"learning_rate": 0.003,
"loss": 3.4808615112304686,
"step": 9500
},
{
"epoch": 0.0209715164208612,
"grad_norm": 0.1322467178106308,
"learning_rate": 0.003,
"loss": 3.4551959228515625,
"step": 9600
},
{
"epoch": 0.021189969716911835,
"grad_norm": 0.13316933810710907,
"learning_rate": 0.003,
"loss": 3.4731900024414064,
"step": 9700
},
{
"epoch": 0.021408423012962474,
"grad_norm": 0.15251101553440094,
"learning_rate": 0.003,
"loss": 3.4879879760742187,
"step": 9800
},
{
"epoch": 0.02162687630901311,
"grad_norm": 0.12604866921901703,
"learning_rate": 0.003,
"loss": 3.440266418457031,
"step": 9900
},
{
"epoch": 0.021845329605063746,
"grad_norm": 0.17345112562179565,
"learning_rate": 0.003,
"loss": 3.406102294921875,
"step": 10000
},
{
"epoch": 0.022063782901114385,
"grad_norm": 0.127431258559227,
"learning_rate": 0.003,
"loss": 3.4641342163085938,
"step": 10100
},
{
"epoch": 0.02228223619716502,
"grad_norm": 0.3970823287963867,
"learning_rate": 0.003,
"loss": 3.4580899047851563,
"step": 10200
},
{
"epoch": 0.02250068949321566,
"grad_norm": 0.12958978116512299,
"learning_rate": 0.003,
"loss": 3.467703857421875,
"step": 10300
},
{
"epoch": 0.022719142789266296,
"grad_norm": 0.12477892637252808,
"learning_rate": 0.003,
"loss": 3.4370376586914064,
"step": 10400
},
{
"epoch": 0.022937596085316935,
"grad_norm": 0.15533806383609772,
"learning_rate": 0.003,
"loss": 3.424752502441406,
"step": 10500
},
{
"epoch": 0.02315604938136757,
"grad_norm": 0.12429562211036682,
"learning_rate": 0.003,
"loss": 3.4563424682617185,
"step": 10600
},
{
"epoch": 0.02337450267741821,
"grad_norm": 0.12259926646947861,
"learning_rate": 0.003,
"loss": 3.452345886230469,
"step": 10700
},
{
"epoch": 0.023592955973468846,
"grad_norm": 0.12801794707775116,
"learning_rate": 0.003,
"loss": 3.428161315917969,
"step": 10800
},
{
"epoch": 0.023811409269519485,
"grad_norm": 0.12190855294466019,
"learning_rate": 0.003,
"loss": 3.4313015747070312,
"step": 10900
},
{
"epoch": 0.02402986256557012,
"grad_norm": 0.15164604783058167,
"learning_rate": 0.003,
"loss": 3.4190289306640627,
"step": 11000
},
{
"epoch": 0.02424831586162076,
"grad_norm": 0.147601917386055,
"learning_rate": 0.003,
"loss": 3.426579895019531,
"step": 11100
},
{
"epoch": 0.024466769157671396,
"grad_norm": 0.2954082787036896,
"learning_rate": 0.003,
"loss": 3.4224453735351563,
"step": 11200
},
{
"epoch": 0.024685222453722035,
"grad_norm": 0.12163235992193222,
"learning_rate": 0.003,
"loss": 3.4398626708984374,
"step": 11300
},
{
"epoch": 0.02490367574977267,
"grad_norm": 0.15871132910251617,
"learning_rate": 0.003,
"loss": 3.245552978515625,
"step": 11400
},
{
"epoch": 0.02512212904582331,
"grad_norm": 0.12936411798000336,
"learning_rate": 0.003,
"loss": 3.463407287597656,
"step": 11500
},
{
"epoch": 0.025340582341873946,
"grad_norm": 0.14481250941753387,
"learning_rate": 0.003,
"loss": 3.4297866821289062,
"step": 11600
},
{
"epoch": 0.025559035637924585,
"grad_norm": 0.1768658310174942,
"learning_rate": 0.003,
"loss": 3.4299252319335936,
"step": 11700
},
{
"epoch": 0.02577748893397522,
"grad_norm": 0.16980133950710297,
"learning_rate": 0.003,
"loss": 3.422637634277344,
"step": 11800
},
{
"epoch": 0.02599594223002586,
"grad_norm": 0.14592786133289337,
"learning_rate": 0.003,
"loss": 3.4101101684570314,
"step": 11900
},
{
"epoch": 0.026214395526076496,
"grad_norm": 0.17400604486465454,
"learning_rate": 0.003,
"loss": 3.4130960083007813,
"step": 12000
},
{
"epoch": 0.026432848822127136,
"grad_norm": 0.15158580243587494,
"learning_rate": 0.003,
"loss": 3.3233709716796875,
"step": 12100
},
{
"epoch": 0.02665130211817777,
"grad_norm": 0.1904173642396927,
"learning_rate": 0.003,
"loss": 3.369674072265625,
"step": 12200
},
{
"epoch": 0.02686975541422841,
"grad_norm": 0.17572258412837982,
"learning_rate": 0.003,
"loss": 3.3876129150390626,
"step": 12300
},
{
"epoch": 0.027088208710279046,
"grad_norm": 0.17185264825820923,
"learning_rate": 0.003,
"loss": 3.4066583251953126,
"step": 12400
},
{
"epoch": 0.027306662006329686,
"grad_norm": 0.1479792445898056,
"learning_rate": 0.003,
"loss": 3.4205966186523438,
"step": 12500
},
{
"epoch": 0.02752511530238032,
"grad_norm": 0.12834446132183075,
"learning_rate": 0.003,
"loss": 3.378619384765625,
"step": 12600
},
{
"epoch": 0.02774356859843096,
"grad_norm": 0.12408050894737244,
"learning_rate": 0.003,
"loss": 3.3828253173828124,
"step": 12700
},
{
"epoch": 0.027962021894481597,
"grad_norm": 0.22868679463863373,
"learning_rate": 0.003,
"loss": 3.381539306640625,
"step": 12800
},
{
"epoch": 0.028180475190532236,
"grad_norm": 0.17433251440525055,
"learning_rate": 0.003,
"loss": 3.4322048950195314,
"step": 12900
},
{
"epoch": 0.02839892848658287,
"grad_norm": 0.16155290603637695,
"learning_rate": 0.003,
"loss": 3.4481887817382812,
"step": 13000
},
{
"epoch": 0.028617381782633507,
"grad_norm": 0.15982672572135925,
"learning_rate": 0.003,
"loss": 3.4119744873046876,
"step": 13100
},
{
"epoch": 0.028835835078684147,
"grad_norm": 0.18386831879615784,
"learning_rate": 0.003,
"loss": 3.3510791015625,
"step": 13200
},
{
"epoch": 0.029054288374734782,
"grad_norm": 0.17137445509433746,
"learning_rate": 0.003,
"loss": 3.393916015625,
"step": 13300
},
{
"epoch": 0.02927274167078542,
"grad_norm": 0.13146288692951202,
"learning_rate": 0.003,
"loss": 3.3667041015625,
"step": 13400
},
{
"epoch": 0.029491194966836057,
"grad_norm": 0.14492550492286682,
"learning_rate": 0.003,
"loss": 3.419930114746094,
"step": 13500
},
{
"epoch": 0.029709648262886697,
"grad_norm": 0.1275530457496643,
"learning_rate": 0.003,
"loss": 3.3596307373046876,
"step": 13600
},
{
"epoch": 0.029928101558937333,
"grad_norm": 0.13028651475906372,
"learning_rate": 0.003,
"loss": 3.4040185546875,
"step": 13700
},
{
"epoch": 0.030146554854987972,
"grad_norm": 0.11924244463443756,
"learning_rate": 0.003,
"loss": 3.3670413208007814,
"step": 13800
},
{
"epoch": 0.030365008151038608,
"grad_norm": 0.14716394245624542,
"learning_rate": 0.003,
"loss": 3.3846261596679685,
"step": 13900
},
{
"epoch": 0.030583461447089247,
"grad_norm": 0.13254448771476746,
"learning_rate": 0.003,
"loss": 3.364213562011719,
"step": 14000
},
{
"epoch": 0.030801914743139883,
"grad_norm": 0.15664011240005493,
"learning_rate": 0.003,
"loss": 3.3782522583007815,
"step": 14100
},
{
"epoch": 0.031020368039190522,
"grad_norm": 0.15272946655750275,
"learning_rate": 0.003,
"loss": 3.3805450439453124,
"step": 14200
},
{
"epoch": 0.031238821335241158,
"grad_norm": 0.1525123417377472,
"learning_rate": 0.003,
"loss": 3.3752081298828127,
"step": 14300
},
{
"epoch": 0.0314572746312918,
"grad_norm": 0.20490288734436035,
"learning_rate": 0.003,
"loss": 3.371563720703125,
"step": 14400
},
{
"epoch": 0.03167572792734243,
"grad_norm": 0.15253432095050812,
"learning_rate": 0.003,
"loss": 3.402916259765625,
"step": 14500
},
{
"epoch": 0.03189418122339307,
"grad_norm": 0.13024139404296875,
"learning_rate": 0.003,
"loss": 3.4280859375,
"step": 14600
},
{
"epoch": 0.03211263451944371,
"grad_norm": 0.1582784652709961,
"learning_rate": 0.003,
"loss": 3.3901861572265624,
"step": 14700
},
{
"epoch": 0.03233108781549435,
"grad_norm": 0.14428076148033142,
"learning_rate": 0.003,
"loss": 3.4218453979492187,
"step": 14800
},
{
"epoch": 0.03254954111154498,
"grad_norm": 0.17758288979530334,
"learning_rate": 0.003,
"loss": 3.414378662109375,
"step": 14900
},
{
"epoch": 0.03276799440759562,
"grad_norm": 0.15624703466892242,
"learning_rate": 0.003,
"loss": 3.402666015625,
"step": 15000
},
{
"epoch": 0.03298644770364626,
"grad_norm": 0.11428515613079071,
"learning_rate": 0.003,
"loss": 3.3660311889648438,
"step": 15100
},
{
"epoch": 0.0332049009996969,
"grad_norm": 0.12380622327327728,
"learning_rate": 0.003,
"loss": 3.373631591796875,
"step": 15200
},
{
"epoch": 0.03342335429574753,
"grad_norm": 0.1253189742565155,
"learning_rate": 0.003,
"loss": 3.4366766357421876,
"step": 15300
},
{
"epoch": 0.03364180759179817,
"grad_norm": 0.16725805401802063,
"learning_rate": 0.003,
"loss": 3.3884176635742187,
"step": 15400
},
{
"epoch": 0.03386026088784881,
"grad_norm": 0.11203224956989288,
"learning_rate": 0.003,
"loss": 3.38418701171875,
"step": 15500
},
{
"epoch": 0.03407871418389945,
"grad_norm": 0.14427132904529572,
"learning_rate": 0.003,
"loss": 3.35841552734375,
"step": 15600
},
{
"epoch": 0.03429716747995008,
"grad_norm": 0.3464087247848511,
"learning_rate": 0.003,
"loss": 3.3959381103515627,
"step": 15700
},
{
"epoch": 0.03451562077600072,
"grad_norm": 0.1291625201702118,
"learning_rate": 0.003,
"loss": 3.318126220703125,
"step": 15800
},
{
"epoch": 0.03473407407205136,
"grad_norm": 0.1512761414051056,
"learning_rate": 0.003,
"loss": 3.4096249389648436,
"step": 15900
},
{
"epoch": 0.034952527368102,
"grad_norm": 0.13020943105220795,
"learning_rate": 0.003,
"loss": 3.394733581542969,
"step": 16000
},
{
"epoch": 0.03517098066415263,
"grad_norm": 0.1642727553844452,
"learning_rate": 0.003,
"loss": 3.387644958496094,
"step": 16100
},
{
"epoch": 0.03538943396020327,
"grad_norm": 0.18252240121364594,
"learning_rate": 0.003,
"loss": 3.348713684082031,
"step": 16200
},
{
"epoch": 0.03560788725625391,
"grad_norm": 0.12791410088539124,
"learning_rate": 0.003,
"loss": 3.3396737670898435,
"step": 16300
},
{
"epoch": 0.03582634055230455,
"grad_norm": 0.16357003152370453,
"learning_rate": 0.003,
"loss": 3.3477932739257814,
"step": 16400
},
{
"epoch": 0.03604479384835518,
"grad_norm": 0.15510688722133636,
"learning_rate": 0.003,
"loss": 3.359996337890625,
"step": 16500
},
{
"epoch": 0.03626324714440582,
"grad_norm": 0.13001827895641327,
"learning_rate": 0.003,
"loss": 3.389567565917969,
"step": 16600
},
{
"epoch": 0.036481700440456455,
"grad_norm": 0.1236167848110199,
"learning_rate": 0.003,
"loss": 3.3922064208984377,
"step": 16700
},
{
"epoch": 0.0367001537365071,
"grad_norm": 0.1220364198088646,
"learning_rate": 0.003,
"loss": 3.36064453125,
"step": 16800
},
{
"epoch": 0.03691860703255773,
"grad_norm": 0.15455572307109833,
"learning_rate": 0.003,
"loss": 3.3853970336914063,
"step": 16900
},
{
"epoch": 0.03713706032860837,
"grad_norm": 0.1555703580379486,
"learning_rate": 0.003,
"loss": 3.3637930297851564,
"step": 17000
},
{
"epoch": 0.037355513624659005,
"grad_norm": 0.12709327042102814,
"learning_rate": 0.003,
"loss": 3.3593380737304686,
"step": 17100
},
{
"epoch": 0.03757396692070965,
"grad_norm": 0.1316322535276413,
"learning_rate": 0.003,
"loss": 3.3686447143554688,
"step": 17200
},
{
"epoch": 0.037792420216760284,
"grad_norm": 0.1177879050374031,
"learning_rate": 0.003,
"loss": 3.395923767089844,
"step": 17300
},
{
"epoch": 0.03801087351281092,
"grad_norm": 0.13159431517124176,
"learning_rate": 0.003,
"loss": 3.3620074462890623,
"step": 17400
},
{
"epoch": 0.038229326808861555,
"grad_norm": 0.18160413205623627,
"learning_rate": 0.003,
"loss": 3.399137268066406,
"step": 17500
},
{
"epoch": 0.0384477801049122,
"grad_norm": 0.14702297747135162,
"learning_rate": 0.003,
"loss": 3.3676507568359373,
"step": 17600
},
{
"epoch": 0.038666233400962834,
"grad_norm": 0.1425740122795105,
"learning_rate": 0.003,
"loss": 3.369291076660156,
"step": 17700
},
{
"epoch": 0.03888468669701347,
"grad_norm": 0.14931966364383698,
"learning_rate": 0.003,
"loss": 3.4102838134765623,
"step": 17800
},
{
"epoch": 0.039103139993064105,
"grad_norm": 0.17999820411205292,
"learning_rate": 0.003,
"loss": 3.379178466796875,
"step": 17900
},
{
"epoch": 0.03932159328911475,
"grad_norm": 0.1291344314813614,
"learning_rate": 0.003,
"loss": 3.355089111328125,
"step": 18000
},
{
"epoch": 0.039540046585165384,
"grad_norm": 0.1446428745985031,
"learning_rate": 0.003,
"loss": 3.3309194946289065,
"step": 18100
},
{
"epoch": 0.03975849988121602,
"grad_norm": 0.14135949313640594,
"learning_rate": 0.003,
"loss": 3.3592868041992188,
"step": 18200
},
{
"epoch": 0.039976953177266655,
"grad_norm": 0.1341768056154251,
"learning_rate": 0.003,
"loss": 3.4569125366210938,
"step": 18300
},
{
"epoch": 0.0401954064733173,
"grad_norm": 0.17552950978279114,
"learning_rate": 0.003,
"loss": 3.3678436279296875,
"step": 18400
},
{
"epoch": 0.040413859769367934,
"grad_norm": 0.13788950443267822,
"learning_rate": 0.003,
"loss": 3.3702731323242188,
"step": 18500
},
{
"epoch": 0.04063231306541857,
"grad_norm": 0.15055698156356812,
"learning_rate": 0.003,
"loss": 3.3611465454101563,
"step": 18600
},
{
"epoch": 0.040850766361469205,
"grad_norm": 0.14934693276882172,
"learning_rate": 0.003,
"loss": 3.3552645874023437,
"step": 18700
},
{
"epoch": 0.04106921965751985,
"grad_norm": 0.14292465150356293,
"learning_rate": 0.003,
"loss": 3.366205139160156,
"step": 18800
},
{
"epoch": 0.041287672953570484,
"grad_norm": 0.18903979659080505,
"learning_rate": 0.003,
"loss": 3.371684265136719,
"step": 18900
},
{
"epoch": 0.04150612624962112,
"grad_norm": 0.24123075604438782,
"learning_rate": 0.003,
"loss": 3.3701324462890625,
"step": 19000
},
{
"epoch": 0.041724579545671756,
"grad_norm": 0.07729237526655197,
"learning_rate": 0.003,
"loss": 3.298961181640625,
"step": 19100
},
{
"epoch": 0.0419430328417224,
"grad_norm": 0.1265183836221695,
"learning_rate": 0.003,
"loss": 3.3053384399414063,
"step": 19200
},
{
"epoch": 0.042161486137773034,
"grad_norm": 0.1463397592306137,
"learning_rate": 0.003,
"loss": 3.268896179199219,
"step": 19300
},
{
"epoch": 0.04237993943382367,
"grad_norm": 0.16664332151412964,
"learning_rate": 0.003,
"loss": 3.298565673828125,
"step": 19400
},
{
"epoch": 0.042598392729874306,
"grad_norm": 0.191655695438385,
"learning_rate": 0.003,
"loss": 3.362577209472656,
"step": 19500
},
{
"epoch": 0.04281684602592495,
"grad_norm": 0.15652531385421753,
"learning_rate": 0.003,
"loss": 3.382750244140625,
"step": 19600
},
{
"epoch": 0.043035299321975584,
"grad_norm": 0.1691717952489853,
"learning_rate": 0.003,
"loss": 3.38729248046875,
"step": 19700
},
{
"epoch": 0.04325375261802622,
"grad_norm": 0.17442384362220764,
"learning_rate": 0.003,
"loss": 3.4193304443359374,
"step": 19800
},
{
"epoch": 0.043472205914076856,
"grad_norm": 0.1332450658082962,
"learning_rate": 0.003,
"loss": 3.392738037109375,
"step": 19900
},
{
"epoch": 0.04369065921012749,
"grad_norm": 0.21232371032238007,
"learning_rate": 0.003,
"loss": 3.302718200683594,
"step": 20000
},
{
"epoch": 0.043909112506178134,
"grad_norm": 0.12638093531131744,
"learning_rate": 0.003,
"loss": 3.3222625732421873,
"step": 20100
},
{
"epoch": 0.04412756580222877,
"grad_norm": 0.14597079157829285,
"learning_rate": 0.003,
"loss": 3.3344183349609375,
"step": 20200
},
{
"epoch": 0.044346019098279406,
"grad_norm": 0.18109019100666046,
"learning_rate": 0.003,
"loss": 3.355440979003906,
"step": 20300
},
{
"epoch": 0.04456447239433004,
"grad_norm": 0.13938353955745697,
"learning_rate": 0.003,
"loss": 3.356192321777344,
"step": 20400
},
{
"epoch": 0.044782925690380684,
"grad_norm": 0.15982763469219208,
"learning_rate": 0.003,
"loss": 3.363128662109375,
"step": 20500
},
{
"epoch": 0.04500137898643132,
"grad_norm": 0.13512779772281647,
"learning_rate": 0.003,
"loss": 3.352191162109375,
"step": 20600
},
{
"epoch": 0.045219832282481956,
"grad_norm": 0.1318831890821457,
"learning_rate": 0.003,
"loss": 3.3415237426757813,
"step": 20700
},
{
"epoch": 0.04543828557853259,
"grad_norm": 0.1379501074552536,
"learning_rate": 0.003,
"loss": 3.31056884765625,
"step": 20800
},
{
"epoch": 0.045656738874583234,
"grad_norm": 0.13282452523708344,
"learning_rate": 0.003,
"loss": 3.3612637329101562,
"step": 20900
},
{
"epoch": 0.04587519217063387,
"grad_norm": 0.16338206827640533,
"learning_rate": 0.003,
"loss": 3.3540554809570313,
"step": 21000
},
{
"epoch": 0.046093645466684506,
"grad_norm": 0.14283308386802673,
"learning_rate": 0.003,
"loss": 2.806424255371094,
"step": 21100
},
{
"epoch": 0.04631209876273514,
"grad_norm": 0.1394539177417755,
"learning_rate": 0.003,
"loss": 3.337768859863281,
"step": 21200
},
{
"epoch": 0.046530552058785785,
"grad_norm": 0.14424312114715576,
"learning_rate": 0.003,
"loss": 3.3319973754882812,
"step": 21300
},
{
"epoch": 0.04674900535483642,
"grad_norm": 0.13849256932735443,
"learning_rate": 0.003,
"loss": 3.357599182128906,
"step": 21400
},
{
"epoch": 0.046967458650887056,
"grad_norm": 0.16135214269161224,
"learning_rate": 0.003,
"loss": 3.3484396362304687,
"step": 21500
},
{
"epoch": 0.04718591194693769,
"grad_norm": 0.12737597525119781,
"learning_rate": 0.003,
"loss": 3.32648681640625,
"step": 21600
},
{
"epoch": 0.047404365242988335,
"grad_norm": 0.1328800916671753,
"learning_rate": 0.003,
"loss": 3.3115444946289063,
"step": 21700
},
{
"epoch": 0.04762281853903897,
"grad_norm": 0.13899770379066467,
"learning_rate": 0.003,
"loss": 3.315568542480469,
"step": 21800
},
{
"epoch": 0.047841271835089606,
"grad_norm": 0.13824427127838135,
"learning_rate": 0.003,
"loss": 3.3840231323242187,
"step": 21900
},
{
"epoch": 0.04805972513114024,
"grad_norm": 0.13134266436100006,
"learning_rate": 0.003,
"loss": 3.332655944824219,
"step": 22000
},
{
"epoch": 0.048278178427190885,
"grad_norm": 0.155122771859169,
"learning_rate": 0.003,
"loss": 3.3235379028320313,
"step": 22100
},
{
"epoch": 0.04849663172324152,
"grad_norm": 0.15229357779026031,
"learning_rate": 0.003,
"loss": 3.3199755859375,
"step": 22200
},
{
"epoch": 0.048715085019292156,
"grad_norm": 0.12670215964317322,
"learning_rate": 0.003,
"loss": 3.316469421386719,
"step": 22300
},
{
"epoch": 0.04893353831534279,
"grad_norm": 0.141511932015419,
"learning_rate": 0.003,
"loss": 3.3063583374023438,
"step": 22400
},
{
"epoch": 0.049151991611393435,
"grad_norm": 0.1661921739578247,
"learning_rate": 0.003,
"loss": 3.3516119384765624,
"step": 22500
},
{
"epoch": 0.04937044490744407,
"grad_norm": 0.158327117562294,
"learning_rate": 0.003,
"loss": 3.31264892578125,
"step": 22600
},
{
"epoch": 0.049588898203494707,
"grad_norm": 0.12664683163166046,
"learning_rate": 0.003,
"loss": 3.3332479858398436,
"step": 22700
},
{
"epoch": 0.04980735149954534,
"grad_norm": 0.16267594695091248,
"learning_rate": 0.003,
"loss": 3.3688790893554685,
"step": 22800
},
{
"epoch": 0.05002580479559598,
"grad_norm": 0.12839975953102112,
"learning_rate": 0.003,
"loss": 3.342285461425781,
"step": 22900
},
{
"epoch": 0.05024425809164662,
"grad_norm": 0.12727583944797516,
"learning_rate": 0.003,
"loss": 3.3483712768554685,
"step": 23000
},
{
"epoch": 0.05046271138769726,
"grad_norm": 0.1626160889863968,
"learning_rate": 0.003,
"loss": 3.258794250488281,
"step": 23100
},
{
"epoch": 0.05068116468374789,
"grad_norm": 0.1446387618780136,
"learning_rate": 0.003,
"loss": 3.360071105957031,
"step": 23200
},
{
"epoch": 0.05089961797979853,
"grad_norm": 0.14391113817691803,
"learning_rate": 0.003,
"loss": 3.3143096923828126,
"step": 23300
},
{
"epoch": 0.05111807127584917,
"grad_norm": 0.1434401422739029,
"learning_rate": 0.003,
"loss": 3.325472106933594,
"step": 23400
},
{
"epoch": 0.05133652457189981,
"grad_norm": 0.13786746561527252,
"learning_rate": 0.003,
"loss": 3.310267333984375,
"step": 23500
},
{
"epoch": 0.05155497786795044,
"grad_norm": 0.13826313614845276,
"learning_rate": 0.003,
"loss": 3.356964111328125,
"step": 23600
},
{
"epoch": 0.05177343116400108,
"grad_norm": 0.12940509617328644,
"learning_rate": 0.003,
"loss": 3.3435748291015623,
"step": 23700
},
{
"epoch": 0.05199188446005172,
"grad_norm": 0.16012531518936157,
"learning_rate": 0.003,
"loss": 3.377527160644531,
"step": 23800
},
{
"epoch": 0.05221033775610236,
"grad_norm": 0.15281037986278534,
"learning_rate": 0.003,
"loss": 3.3529147338867187,
"step": 23900
},
{
"epoch": 0.05242879105215299,
"grad_norm": 0.12364033609628677,
"learning_rate": 0.003,
"loss": 3.3162277221679686,
"step": 24000
},
{
"epoch": 0.05264724434820363,
"grad_norm": 0.13337066769599915,
"learning_rate": 0.003,
"loss": 3.332524719238281,
"step": 24100
},
{
"epoch": 0.05286569764425427,
"grad_norm": 0.12411975860595703,
"learning_rate": 0.003,
"loss": 3.3512115478515625,
"step": 24200
},
{
"epoch": 0.05308415094030491,
"grad_norm": 0.16036230325698853,
"learning_rate": 0.003,
"loss": 3.3250997924804686,
"step": 24300
},
{
"epoch": 0.05330260423635554,
"grad_norm": 0.22030843794345856,
"learning_rate": 0.003,
"loss": 3.304237060546875,
"step": 24400
},
{
"epoch": 0.05352105753240618,
"grad_norm": 0.25381991267204285,
"learning_rate": 0.003,
"loss": 3.34534423828125,
"step": 24500
},
{
"epoch": 0.05373951082845682,
"grad_norm": 0.16389718651771545,
"learning_rate": 0.003,
"loss": 3.348591613769531,
"step": 24600
},
{
"epoch": 0.05395796412450746,
"grad_norm": 0.14573702216148376,
"learning_rate": 0.003,
"loss": 3.3080172729492188,
"step": 24700
},
{
"epoch": 0.05417641742055809,
"grad_norm": 0.1497870832681656,
"learning_rate": 0.003,
"loss": 3.3553036499023436,
"step": 24800
},
{
"epoch": 0.05439487071660873,
"grad_norm": 0.13177230954170227,
"learning_rate": 0.003,
"loss": 3.2844012451171873,
"step": 24900
},
{
"epoch": 0.05461332401265937,
"grad_norm": 0.15044496953487396,
"learning_rate": 0.003,
"loss": 3.3339718627929686,
"step": 25000
},
{
"epoch": 0.05483177730871001,
"grad_norm": 0.16820241510868073,
"learning_rate": 0.003,
"loss": 3.312646484375,
"step": 25100
},
{
"epoch": 0.05505023060476064,
"grad_norm": 0.13994398713111877,
"learning_rate": 0.003,
"loss": 3.3246527099609375,
"step": 25200
},
{
"epoch": 0.05526868390081128,
"grad_norm": 0.13690656423568726,
"learning_rate": 0.003,
"loss": 3.3303009033203126,
"step": 25300
},
{
"epoch": 0.05548713719686192,
"grad_norm": 0.1266283243894577,
"learning_rate": 0.003,
"loss": 3.3370904541015625,
"step": 25400
},
{
"epoch": 0.05570559049291256,
"grad_norm": 0.2835458815097809,
"learning_rate": 0.003,
"loss": 3.2755780029296875,
"step": 25500
},
{
"epoch": 0.05592404378896319,
"grad_norm": 0.1243748813867569,
"learning_rate": 0.003,
"loss": 3.31575927734375,
"step": 25600
},
{
"epoch": 0.05614249708501383,
"grad_norm": 0.13248547911643982,
"learning_rate": 0.003,
"loss": 3.273077087402344,
"step": 25700
},
{
"epoch": 0.05636095038106447,
"grad_norm": 0.1573888212442398,
"learning_rate": 0.003,
"loss": 3.329945373535156,
"step": 25800
},
{
"epoch": 0.05657940367711511,
"grad_norm": 0.1859431117773056,
"learning_rate": 0.003,
"loss": 3.3422332763671876,
"step": 25900
},
{
"epoch": 0.05679785697316574,
"grad_norm": 0.14315782487392426,
"learning_rate": 0.003,
"loss": 3.321678466796875,
"step": 26000
},
{
"epoch": 0.05701631026921638,
"grad_norm": 0.16119340062141418,
"learning_rate": 0.003,
"loss": 3.28656982421875,
"step": 26100
},
{
"epoch": 0.057234763565267015,
"grad_norm": 0.169467493891716,
"learning_rate": 0.003,
"loss": 3.3227874755859377,
"step": 26200
},
{
"epoch": 0.05745321686131766,
"grad_norm": 0.13390371203422546,
"learning_rate": 0.003,
"loss": 3.3195626831054685,
"step": 26300
},
{
"epoch": 0.05767167015736829,
"grad_norm": 0.2256622463464737,
"learning_rate": 0.003,
"loss": 3.3732449340820314,
"step": 26400
},
{
"epoch": 0.05789012345341893,
"grad_norm": 0.14236938953399658,
"learning_rate": 0.003,
"loss": 3.315290832519531,
"step": 26500
},
{
"epoch": 0.058108576749469565,
"grad_norm": 0.13468952476978302,
"learning_rate": 0.003,
"loss": 3.2961331176757813,
"step": 26600
},
{
"epoch": 0.05832703004552021,
"grad_norm": 0.13163988292217255,
"learning_rate": 0.003,
"loss": 3.3402911376953126,
"step": 26700
},
{
"epoch": 0.05854548334157084,
"grad_norm": 0.16410379111766815,
"learning_rate": 0.003,
"loss": 3.3237750244140627,
"step": 26800
},
{
"epoch": 0.05876393663762148,
"grad_norm": 0.17589592933654785,
"learning_rate": 0.003,
"loss": 3.3025152587890627,
"step": 26900
},
{
"epoch": 0.058982389933672115,
"grad_norm": 0.12554213404655457,
"learning_rate": 0.003,
"loss": 3.329324035644531,
"step": 27000
},
{
"epoch": 0.05920084322972276,
"grad_norm": 0.14356501400470734,
"learning_rate": 0.003,
"loss": 3.3095223999023435,
"step": 27100
},
{
"epoch": 0.059419296525773393,
"grad_norm": 0.20642130076885223,
"learning_rate": 0.003,
"loss": 3.16523681640625,
"step": 27200
},
{
"epoch": 0.05963774982182403,
"grad_norm": 0.12734311819076538,
"learning_rate": 0.003,
"loss": 3.3130838012695314,
"step": 27300
},
{
"epoch": 0.059856203117874665,
"grad_norm": 0.16035129129886627,
"learning_rate": 0.003,
"loss": 3.3009878540039064,
"step": 27400
},
{
"epoch": 0.06007465641392531,
"grad_norm": 0.1451551616191864,
"learning_rate": 0.003,
"loss": 3.3051492309570314,
"step": 27500
},
{
"epoch": 0.060293109709975944,
"grad_norm": 0.12821145355701447,
"learning_rate": 0.003,
"loss": 3.32,
"step": 27600
},
{
"epoch": 0.06051156300602658,
"grad_norm": 0.1578887552022934,
"learning_rate": 0.003,
"loss": 3.2693597412109376,
"step": 27700
},
{
"epoch": 0.060730016302077215,
"grad_norm": 0.1403341293334961,
"learning_rate": 0.003,
"loss": 3.285400085449219,
"step": 27800
},
{
"epoch": 0.06094846959812786,
"grad_norm": 0.1922478973865509,
"learning_rate": 0.003,
"loss": 3.3167987060546875,
"step": 27900
},
{
"epoch": 0.061166922894178494,
"grad_norm": 0.12697449326515198,
"learning_rate": 0.003,
"loss": 3.326105651855469,
"step": 28000
},
{
"epoch": 0.06138537619022913,
"grad_norm": 0.21215832233428955,
"learning_rate": 0.003,
"loss": 3.309931640625,
"step": 28100
},
{
"epoch": 0.061603829486279765,
"grad_norm": 0.13175860047340393,
"learning_rate": 0.003,
"loss": 3.2714743041992187,
"step": 28200
},
{
"epoch": 0.06182228278233041,
"grad_norm": 0.12134334444999695,
"learning_rate": 0.003,
"loss": 3.3450942993164063,
"step": 28300
},
{
"epoch": 0.062040736078381044,
"grad_norm": 0.16979266703128815,
"learning_rate": 0.003,
"loss": 3.277815856933594,
"step": 28400
},
{
"epoch": 0.06225918937443168,
"grad_norm": 0.13171924650669098,
"learning_rate": 0.003,
"loss": 3.2926800537109373,
"step": 28500
},
{
"epoch": 0.062477642670482315,
"grad_norm": 0.13504329323768616,
"learning_rate": 0.003,
"loss": 3.304072265625,
"step": 28600
},
{
"epoch": 0.06269609596653296,
"grad_norm": 0.1422707885503769,
"learning_rate": 0.003,
"loss": 3.2862860107421876,
"step": 28700
},
{
"epoch": 0.0629145492625836,
"grad_norm": 0.17020823061466217,
"learning_rate": 0.003,
"loss": 3.3183663940429686,
"step": 28800
},
{
"epoch": 0.06313300255863423,
"grad_norm": 0.7097678780555725,
"learning_rate": 0.003,
"loss": 3.295379333496094,
"step": 28900
},
{
"epoch": 0.06335145585468487,
"grad_norm": 0.14952735602855682,
"learning_rate": 0.003,
"loss": 3.2717132568359375,
"step": 29000
},
{
"epoch": 0.0635699091507355,
"grad_norm": 0.1692362278699875,
"learning_rate": 0.003,
"loss": 3.3412515258789064,
"step": 29100
},
{
"epoch": 0.06378836244678614,
"grad_norm": 0.15614360570907593,
"learning_rate": 0.003,
"loss": 3.3063565063476563,
"step": 29200
},
{
"epoch": 0.06400681574283679,
"grad_norm": 0.20806823670864105,
"learning_rate": 0.003,
"loss": 3.2843331909179687,
"step": 29300
},
{
"epoch": 0.06422526903888742,
"grad_norm": 0.23479899764060974,
"learning_rate": 0.003,
"loss": 3.2819674682617186,
"step": 29400
},
{
"epoch": 0.06444372233493806,
"grad_norm": 0.15177583694458008,
"learning_rate": 0.003,
"loss": 3.3203561401367185,
"step": 29500
},
{
"epoch": 0.0646621756309887,
"grad_norm": 0.1472543627023697,
"learning_rate": 0.003,
"loss": 3.3170175170898437,
"step": 29600
},
{
"epoch": 0.06488062892703933,
"grad_norm": 0.13402847945690155,
"learning_rate": 0.003,
"loss": 3.3249472045898436,
"step": 29700
},
{
"epoch": 0.06509908222308997,
"grad_norm": 0.1574559062719345,
"learning_rate": 0.003,
"loss": 3.3021722412109376,
"step": 29800
},
{
"epoch": 0.0653175355191406,
"grad_norm": 0.12917636334896088,
"learning_rate": 0.003,
"loss": 3.2708474731445314,
"step": 29900
},
{
"epoch": 0.06553598881519124,
"grad_norm": 0.19614313542842865,
"learning_rate": 0.003,
"loss": 3.350285949707031,
"step": 30000
},
{
"epoch": 0.06575444211124187,
"grad_norm": 0.21345964074134827,
"learning_rate": 0.003,
"loss": 3.2905197143554688,
"step": 30100
},
{
"epoch": 0.06597289540729252,
"grad_norm": 0.13310234248638153,
"learning_rate": 0.003,
"loss": 3.14111572265625,
"step": 30200
},
{
"epoch": 0.06619134870334316,
"grad_norm": 0.1672133207321167,
"learning_rate": 0.003,
"loss": 3.292821960449219,
"step": 30300
},
{
"epoch": 0.0664098019993938,
"grad_norm": 0.15199998021125793,
"learning_rate": 0.003,
"loss": 3.281614990234375,
"step": 30400
},
{
"epoch": 0.06662825529544443,
"grad_norm": 0.17468766868114471,
"learning_rate": 0.003,
"loss": 3.28716552734375,
"step": 30500
},
{
"epoch": 0.06684670859149507,
"grad_norm": 0.13380374014377594,
"learning_rate": 0.003,
"loss": 3.3208758544921877,
"step": 30600
},
{
"epoch": 0.0670651618875457,
"grad_norm": 0.12906578183174133,
"learning_rate": 0.003,
"loss": 3.2482403564453124,
"step": 30700
},
{
"epoch": 0.06728361518359634,
"grad_norm": 0.16773296892642975,
"learning_rate": 0.003,
"loss": 3.2957916259765625,
"step": 30800
},
{
"epoch": 0.06750206847964697,
"grad_norm": 0.1382984071969986,
"learning_rate": 0.003,
"loss": 3.2812515258789063,
"step": 30900
},
{
"epoch": 0.06772052177569762,
"grad_norm": 0.12064798176288605,
"learning_rate": 0.003,
"loss": 3.292807312011719,
"step": 31000
},
{
"epoch": 0.06793897507174826,
"grad_norm": 0.14423973858356476,
"learning_rate": 0.003,
"loss": 3.1367059326171876,
"step": 31100
},
{
"epoch": 0.0681574283677989,
"grad_norm": 0.17670339345932007,
"learning_rate": 0.003,
"loss": 3.3249557495117186,
"step": 31200
},
{
"epoch": 0.06837588166384953,
"grad_norm": 0.13148869574069977,
"learning_rate": 0.003,
"loss": 3.323037414550781,
"step": 31300
},
{
"epoch": 0.06859433495990017,
"grad_norm": 0.25557222962379456,
"learning_rate": 0.003,
"loss": 3.2895745849609375,
"step": 31400
},
{
"epoch": 0.0688127882559508,
"grad_norm": 0.1451721489429474,
"learning_rate": 0.003,
"loss": 3.30470703125,
"step": 31500
},
{
"epoch": 0.06903124155200144,
"grad_norm": 0.14752569794654846,
"learning_rate": 0.003,
"loss": 3.2685122680664063,
"step": 31600
},
{
"epoch": 0.06924969484805207,
"grad_norm": 0.12468330562114716,
"learning_rate": 0.003,
"loss": 3.2936700439453124,
"step": 31700
},
{
"epoch": 0.06946814814410272,
"grad_norm": 0.16558866202831268,
"learning_rate": 0.003,
"loss": 3.3152020263671873,
"step": 31800
},
{
"epoch": 0.06968660144015336,
"grad_norm": 0.1636384129524231,
"learning_rate": 0.003,
"loss": 3.311054992675781,
"step": 31900
},
{
"epoch": 0.069905054736204,
"grad_norm": 0.13477930426597595,
"learning_rate": 0.003,
"loss": 3.2879302978515623,
"step": 32000
},
{
"epoch": 0.07012350803225463,
"grad_norm": 0.13240183889865875,
"learning_rate": 0.003,
"loss": 3.337416687011719,
"step": 32100
},
{
"epoch": 0.07034196132830527,
"grad_norm": 0.1719566434621811,
"learning_rate": 0.003,
"loss": 3.2947512817382814,
"step": 32200
},
{
"epoch": 0.0705604146243559,
"grad_norm": 0.14052827656269073,
"learning_rate": 0.003,
"loss": 3.282734680175781,
"step": 32300
},
{
"epoch": 0.07077886792040654,
"grad_norm": 0.13085907697677612,
"learning_rate": 0.003,
"loss": 3.297210693359375,
"step": 32400
},
{
"epoch": 0.07099732121645717,
"grad_norm": 0.20025353133678436,
"learning_rate": 0.003,
"loss": 3.2842779541015625,
"step": 32500
},
{
"epoch": 0.07121577451250782,
"grad_norm": 0.1402798742055893,
"learning_rate": 0.003,
"loss": 3.298916931152344,
"step": 32600
},
{
"epoch": 0.07143422780855846,
"grad_norm": 0.14476116001605988,
"learning_rate": 0.003,
"loss": 3.2706668090820314,
"step": 32700
},
{
"epoch": 0.0716526811046091,
"grad_norm": 0.13569767773151398,
"learning_rate": 0.003,
"loss": 3.2827581787109374,
"step": 32800
},
{
"epoch": 0.07187113440065973,
"grad_norm": 0.15818500518798828,
"learning_rate": 0.003,
"loss": 3.2876434326171875,
"step": 32900
},
{
"epoch": 0.07208958769671037,
"grad_norm": 0.1276344656944275,
"learning_rate": 0.003,
"loss": 3.3087249755859376,
"step": 33000
},
{
"epoch": 0.072308040992761,
"grad_norm": 0.13138215243816376,
"learning_rate": 0.003,
"loss": 3.2812591552734376,
"step": 33100
},
{
"epoch": 0.07252649428881164,
"grad_norm": 0.18934592604637146,
"learning_rate": 0.003,
"loss": 3.30925537109375,
"step": 33200
},
{
"epoch": 0.07274494758486227,
"grad_norm": 0.14984296262264252,
"learning_rate": 0.003,
"loss": 3.2924786376953126,
"step": 33300
},
{
"epoch": 0.07296340088091291,
"grad_norm": 0.16432835161685944,
"learning_rate": 0.003,
"loss": 3.279039306640625,
"step": 33400
},
{
"epoch": 0.07318185417696356,
"grad_norm": 0.131127268075943,
"learning_rate": 0.003,
"loss": 3.3125900268554687,
"step": 33500
},
{
"epoch": 0.0734003074730142,
"grad_norm": 0.15136845409870148,
"learning_rate": 0.003,
"loss": 3.3073272705078125,
"step": 33600
},
{
"epoch": 0.07361876076906483,
"grad_norm": 0.1228429302573204,
"learning_rate": 0.003,
"loss": 3.1849093627929688,
"step": 33700
},
{
"epoch": 0.07383721406511547,
"grad_norm": 0.12898238003253937,
"learning_rate": 0.003,
"loss": 3.2927169799804688,
"step": 33800
},
{
"epoch": 0.0740556673611661,
"grad_norm": 0.3990197777748108,
"learning_rate": 0.003,
"loss": 3.3054901123046876,
"step": 33900
},
{
"epoch": 0.07427412065721674,
"grad_norm": 0.1889471411705017,
"learning_rate": 0.003,
"loss": 3.2711798095703126,
"step": 34000
},
{
"epoch": 0.07449257395326737,
"grad_norm": 0.32053592801094055,
"learning_rate": 0.003,
"loss": 3.277731628417969,
"step": 34100
},
{
"epoch": 0.07471102724931801,
"grad_norm": 0.12476284056901932,
"learning_rate": 0.003,
"loss": 3.3241058349609376,
"step": 34200
},
{
"epoch": 0.07492948054536866,
"grad_norm": 0.1345445066690445,
"learning_rate": 0.003,
"loss": 3.235545349121094,
"step": 34300
},
{
"epoch": 0.0751479338414193,
"grad_norm": 0.17160938680171967,
"learning_rate": 0.003,
"loss": 3.260941162109375,
"step": 34400
},
{
"epoch": 0.07536638713746993,
"grad_norm": 0.15105551481246948,
"learning_rate": 0.003,
"loss": 3.2895513916015626,
"step": 34500
},
{
"epoch": 0.07558484043352057,
"grad_norm": 0.13841792941093445,
"learning_rate": 0.003,
"loss": 3.3118740844726564,
"step": 34600
},
{
"epoch": 0.0758032937295712,
"grad_norm": 0.13309921324253082,
"learning_rate": 0.003,
"loss": 3.2925543212890624,
"step": 34700
},
{
"epoch": 0.07602174702562184,
"grad_norm": 0.13050492107868195,
"learning_rate": 0.003,
"loss": 3.2816531372070314,
"step": 34800
},
{
"epoch": 0.07624020032167247,
"grad_norm": 0.6023688316345215,
"learning_rate": 0.003,
"loss": 3.3055108642578124,
"step": 34900
},
{
"epoch": 0.07645865361772311,
"grad_norm": 0.1407451182603836,
"learning_rate": 0.003,
"loss": 3.34151611328125,
"step": 35000
},
{
"epoch": 0.07667710691377376,
"grad_norm": 0.1305139809846878,
"learning_rate": 0.003,
"loss": 3.2856158447265624,
"step": 35100
},
{
"epoch": 0.0768955602098244,
"grad_norm": 0.1610284149646759,
"learning_rate": 0.003,
"loss": 3.2734844970703123,
"step": 35200
},
{
"epoch": 0.07711401350587503,
"grad_norm": 0.13967597484588623,
"learning_rate": 0.003,
"loss": 3.256042175292969,
"step": 35300
},
{
"epoch": 0.07733246680192567,
"grad_norm": 0.19864661991596222,
"learning_rate": 0.003,
"loss": 3.324942321777344,
"step": 35400
},
{
"epoch": 0.0775509200979763,
"grad_norm": 0.16039150953292847,
"learning_rate": 0.003,
"loss": 3.2831130981445313,
"step": 35500
},
{
"epoch": 0.07776937339402694,
"grad_norm": 0.1747000366449356,
"learning_rate": 0.003,
"loss": 3.307495422363281,
"step": 35600
},
{
"epoch": 0.07798782669007757,
"grad_norm": 0.20273174345493317,
"learning_rate": 0.003,
"loss": 3.3177590942382813,
"step": 35700
},
{
"epoch": 0.07820627998612821,
"grad_norm": 0.1993030309677124,
"learning_rate": 0.003,
"loss": 3.304837341308594,
"step": 35800
},
{
"epoch": 0.07842473328217886,
"grad_norm": 0.17900240421295166,
"learning_rate": 0.003,
"loss": 3.235675048828125,
"step": 35900
},
{
"epoch": 0.0786431865782295,
"grad_norm": 0.14110247790813446,
"learning_rate": 0.003,
"loss": 3.19402587890625,
"step": 36000
},
{
"epoch": 0.07886163987428013,
"grad_norm": 0.2041473090648651,
"learning_rate": 0.003,
"loss": 3.3216961669921874,
"step": 36100
},
{
"epoch": 0.07908009317033077,
"grad_norm": 0.19810523092746735,
"learning_rate": 0.003,
"loss": 3.2388531494140627,
"step": 36200
},
{
"epoch": 0.0792985464663814,
"grad_norm": 0.12772241234779358,
"learning_rate": 0.003,
"loss": 3.2910110473632814,
"step": 36300
},
{
"epoch": 0.07951699976243204,
"grad_norm": 0.15931345522403717,
"learning_rate": 0.003,
"loss": 3.2836883544921873,
"step": 36400
},
{
"epoch": 0.07973545305848267,
"grad_norm": 0.13564768433570862,
"learning_rate": 0.003,
"loss": 3.3310623168945312,
"step": 36500
},
{
"epoch": 0.07995390635453331,
"grad_norm": 0.15236645936965942,
"learning_rate": 0.003,
"loss": 3.273614501953125,
"step": 36600
},
{
"epoch": 0.08017235965058395,
"grad_norm": 0.14576539397239685,
"learning_rate": 0.003,
"loss": 3.3081988525390624,
"step": 36700
},
{
"epoch": 0.0803908129466346,
"grad_norm": 0.1572989672422409,
"learning_rate": 0.003,
"loss": 3.251156005859375,
"step": 36800
},
{
"epoch": 0.08060926624268523,
"grad_norm": 0.15005140006542206,
"learning_rate": 0.003,
"loss": 3.2585882568359374,
"step": 36900
},
{
"epoch": 0.08082771953873587,
"grad_norm": 0.15507107973098755,
"learning_rate": 0.003,
"loss": 3.266631774902344,
"step": 37000
},
{
"epoch": 0.0810461728347865,
"grad_norm": 0.17641592025756836,
"learning_rate": 0.003,
"loss": 3.239948425292969,
"step": 37100
},
{
"epoch": 0.08126462613083714,
"grad_norm": 0.12613926827907562,
"learning_rate": 0.003,
"loss": 3.3062588500976564,
"step": 37200
},
{
"epoch": 0.08148307942688778,
"grad_norm": 0.2538757920265198,
"learning_rate": 0.003,
"loss": 3.278668212890625,
"step": 37300
},
{
"epoch": 0.08170153272293841,
"grad_norm": 0.16014288365840912,
"learning_rate": 0.003,
"loss": 3.23365966796875,
"step": 37400
},
{
"epoch": 0.08191998601898905,
"grad_norm": 0.14398036897182465,
"learning_rate": 0.003,
"loss": 3.3074264526367188,
"step": 37500
},
{
"epoch": 0.0821384393150397,
"grad_norm": 0.14527015388011932,
"learning_rate": 0.003,
"loss": 3.260564270019531,
"step": 37600
},
{
"epoch": 0.08235689261109033,
"grad_norm": 0.13640612363815308,
"learning_rate": 0.003,
"loss": 3.2675531005859373,
"step": 37700
},
{
"epoch": 0.08257534590714097,
"grad_norm": 0.1483987420797348,
"learning_rate": 0.003,
"loss": 3.241988525390625,
"step": 37800
},
{
"epoch": 0.0827937992031916,
"grad_norm": 0.14941224455833435,
"learning_rate": 0.003,
"loss": 3.3164938354492186,
"step": 37900
},
{
"epoch": 0.08301225249924224,
"grad_norm": 0.17315177619457245,
"learning_rate": 0.003,
"loss": 3.316810302734375,
"step": 38000
},
{
"epoch": 0.08323070579529288,
"grad_norm": 0.17823275923728943,
"learning_rate": 0.003,
"loss": 3.249942626953125,
"step": 38100
},
{
"epoch": 0.08344915909134351,
"grad_norm": 0.20452110469341278,
"learning_rate": 0.003,
"loss": 3.2030743408203124,
"step": 38200
},
{
"epoch": 0.08366761238739415,
"grad_norm": 0.13244900107383728,
"learning_rate": 0.003,
"loss": 3.272771911621094,
"step": 38300
},
{
"epoch": 0.0838860656834448,
"grad_norm": 0.14962248504161835,
"learning_rate": 0.003,
"loss": 3.3190447998046877,
"step": 38400
},
{
"epoch": 0.08410451897949543,
"grad_norm": 0.13936245441436768,
"learning_rate": 0.003,
"loss": 3.2822772216796876,
"step": 38500
},
{
"epoch": 0.08432297227554607,
"grad_norm": 0.17270345985889435,
"learning_rate": 0.003,
"loss": 3.3033078002929686,
"step": 38600
},
{
"epoch": 0.0845414255715967,
"grad_norm": 0.17023159563541412,
"learning_rate": 0.003,
"loss": 3.2737576293945314,
"step": 38700
},
{
"epoch": 0.08475987886764734,
"grad_norm": 0.1845265030860901,
"learning_rate": 0.003,
"loss": 3.2791680908203125,
"step": 38800
},
{
"epoch": 0.08497833216369798,
"grad_norm": 0.13376891613006592,
"learning_rate": 0.003,
"loss": 3.2831585693359373,
"step": 38900
},
{
"epoch": 0.08519678545974861,
"grad_norm": 0.1504724621772766,
"learning_rate": 0.003,
"loss": 3.3097457885742188,
"step": 39000
},
{
"epoch": 0.08541523875579925,
"grad_norm": 0.137648805975914,
"learning_rate": 0.003,
"loss": 3.2856106567382812,
"step": 39100
},
{
"epoch": 0.0856336920518499,
"grad_norm": 0.1618202179670334,
"learning_rate": 0.003,
"loss": 3.267169189453125,
"step": 39200
},
{
"epoch": 0.08585214534790053,
"grad_norm": 0.19786691665649414,
"learning_rate": 0.003,
"loss": 3.2910052490234376,
"step": 39300
},
{
"epoch": 0.08607059864395117,
"grad_norm": 0.15184424817562103,
"learning_rate": 0.003,
"loss": 3.317454833984375,
"step": 39400
},
{
"epoch": 0.0862890519400018,
"grad_norm": 9.05631160736084,
"learning_rate": 0.003,
"loss": 3.2767315673828126,
"step": 39500
},
{
"epoch": 0.08650750523605244,
"grad_norm": 0.15465736389160156,
"learning_rate": 0.003,
"loss": 2.794194641113281,
"step": 39600
},
{
"epoch": 0.08672595853210308,
"grad_norm": 0.16257505118846893,
"learning_rate": 0.003,
"loss": 3.2999783325195313,
"step": 39700
},
{
"epoch": 0.08694441182815371,
"grad_norm": 0.15374092757701874,
"learning_rate": 0.003,
"loss": 3.267034606933594,
"step": 39800
},
{
"epoch": 0.08716286512420435,
"grad_norm": 0.18448083102703094,
"learning_rate": 0.003,
"loss": 3.2641702270507813,
"step": 39900
},
{
"epoch": 0.08738131842025498,
"grad_norm": 0.12463884055614471,
"learning_rate": 0.003,
"loss": 3.284293518066406,
"step": 40000
},
{
"epoch": 0.08759977171630563,
"grad_norm": 0.2346997857093811,
"learning_rate": 0.003,
"loss": 3.3378842163085936,
"step": 40100
},
{
"epoch": 0.08781822501235627,
"grad_norm": 0.16173656284809113,
"learning_rate": 0.003,
"loss": 3.28466064453125,
"step": 40200
},
{
"epoch": 0.0880366783084069,
"grad_norm": 0.15595772862434387,
"learning_rate": 0.003,
"loss": 3.260408630371094,
"step": 40300
},
{
"epoch": 0.08825513160445754,
"grad_norm": 0.12999649345874786,
"learning_rate": 0.003,
"loss": 3.301771545410156,
"step": 40400
},
{
"epoch": 0.08847358490050818,
"grad_norm": 0.12810906767845154,
"learning_rate": 0.003,
"loss": 3.273116455078125,
"step": 40500
},
{
"epoch": 0.08869203819655881,
"grad_norm": 0.18569885194301605,
"learning_rate": 0.003,
"loss": 3.2856179809570314,
"step": 40600
},
{
"epoch": 0.08891049149260945,
"grad_norm": 0.18256127834320068,
"learning_rate": 0.003,
"loss": 3.24281982421875,
"step": 40700
},
{
"epoch": 0.08912894478866008,
"grad_norm": 0.1585642695426941,
"learning_rate": 0.003,
"loss": 3.2112246704101564,
"step": 40800
},
{
"epoch": 0.08934739808471073,
"grad_norm": 0.17732886970043182,
"learning_rate": 0.003,
"loss": 3.2847430419921877,
"step": 40900
},
{
"epoch": 0.08956585138076137,
"grad_norm": 0.13492503762245178,
"learning_rate": 0.003,
"loss": 3.312560729980469,
"step": 41000
},
{
"epoch": 0.089784304676812,
"grad_norm": 0.16829057037830353,
"learning_rate": 0.003,
"loss": 3.2641729736328124,
"step": 41100
},
{
"epoch": 0.09000275797286264,
"grad_norm": 0.18808752298355103,
"learning_rate": 0.003,
"loss": 3.2674325561523436,
"step": 41200
},
{
"epoch": 0.09022121126891328,
"grad_norm": 0.15025244653224945,
"learning_rate": 0.003,
"loss": 3.2771319580078124,
"step": 41300
},
{
"epoch": 0.09043966456496391,
"grad_norm": 0.15831509232521057,
"learning_rate": 0.003,
"loss": 3.2885198974609375,
"step": 41400
},
{
"epoch": 0.09065811786101455,
"grad_norm": 0.1266457885503769,
"learning_rate": 0.003,
"loss": 3.268582763671875,
"step": 41500
},
{
"epoch": 0.09087657115706518,
"grad_norm": 0.18740995228290558,
"learning_rate": 0.003,
"loss": 3.2911370849609374,
"step": 41600
},
{
"epoch": 0.09109502445311583,
"grad_norm": 0.1677059382200241,
"learning_rate": 0.003,
"loss": 3.32934814453125,
"step": 41700
},
{
"epoch": 0.09131347774916647,
"grad_norm": 0.1806642860174179,
"learning_rate": 0.003,
"loss": 3.293853759765625,
"step": 41800
},
{
"epoch": 0.0915319310452171,
"grad_norm": 0.20180915296077728,
"learning_rate": 0.003,
"loss": 3.2697601318359375,
"step": 41900
},
{
"epoch": 0.09175038434126774,
"grad_norm": 0.13965576887130737,
"learning_rate": 0.003,
"loss": 3.28150146484375,
"step": 42000
},
{
"epoch": 0.09196883763731838,
"grad_norm": 0.15903182327747345,
"learning_rate": 0.003,
"loss": 3.2683795166015623,
"step": 42100
},
{
"epoch": 0.09218729093336901,
"grad_norm": 0.1410466581583023,
"learning_rate": 0.003,
"loss": 3.264806823730469,
"step": 42200
},
{
"epoch": 0.09240574422941965,
"grad_norm": 0.1510380655527115,
"learning_rate": 0.003,
"loss": 3.2732772827148438,
"step": 42300
},
{
"epoch": 0.09262419752547028,
"grad_norm": 0.13428038358688354,
"learning_rate": 0.003,
"loss": 3.2841482543945313,
"step": 42400
},
{
"epoch": 0.09284265082152092,
"grad_norm": 0.17533539235591888,
"learning_rate": 0.003,
"loss": 3.2517266845703126,
"step": 42500
},
{
"epoch": 0.09306110411757157,
"grad_norm": 0.1411064714193344,
"learning_rate": 0.003,
"loss": 3.282726135253906,
"step": 42600
},
{
"epoch": 0.0932795574136222,
"grad_norm": 0.19230180978775024,
"learning_rate": 0.003,
"loss": 3.2695498657226563,
"step": 42700
},
{
"epoch": 0.09349801070967284,
"grad_norm": 0.16757774353027344,
"learning_rate": 0.003,
"loss": 3.231575622558594,
"step": 42800
},
{
"epoch": 0.09371646400572348,
"grad_norm": 0.17299550771713257,
"learning_rate": 0.003,
"loss": 3.28582763671875,
"step": 42900
},
{
"epoch": 0.09393491730177411,
"grad_norm": 0.17058102786540985,
"learning_rate": 0.003,
"loss": 3.252261047363281,
"step": 43000
},
{
"epoch": 0.09415337059782475,
"grad_norm": 0.17918632924556732,
"learning_rate": 0.003,
"loss": 3.276212158203125,
"step": 43100
},
{
"epoch": 0.09437182389387538,
"grad_norm": 0.16316772997379303,
"learning_rate": 0.003,
"loss": 3.2334857177734375,
"step": 43200
},
{
"epoch": 0.09459027718992602,
"grad_norm": 0.1895044595003128,
"learning_rate": 0.003,
"loss": 3.2556378173828127,
"step": 43300
},
{
"epoch": 0.09480873048597667,
"grad_norm": 0.15873974561691284,
"learning_rate": 0.003,
"loss": 3.2901712036132813,
"step": 43400
},
{
"epoch": 0.0950271837820273,
"grad_norm": 0.17835567891597748,
"learning_rate": 0.003,
"loss": 3.264893798828125,
"step": 43500
},
{
"epoch": 0.09524563707807794,
"grad_norm": 0.15636888146400452,
"learning_rate": 0.003,
"loss": 3.2413174438476564,
"step": 43600
},
{
"epoch": 0.09546409037412858,
"grad_norm": 0.14264269173145294,
"learning_rate": 0.003,
"loss": 3.268505859375,
"step": 43700
},
{
"epoch": 0.09568254367017921,
"grad_norm": 0.15490996837615967,
"learning_rate": 0.003,
"loss": 3.235081787109375,
"step": 43800
},
{
"epoch": 0.09590099696622985,
"grad_norm": 0.13821007311344147,
"learning_rate": 0.003,
"loss": 3.296794128417969,
"step": 43900
},
{
"epoch": 0.09611945026228048,
"grad_norm": 0.16242928802967072,
"learning_rate": 0.003,
"loss": 3.278422546386719,
"step": 44000
},
{
"epoch": 0.09633790355833112,
"grad_norm": 0.15828748047351837,
"learning_rate": 0.003,
"loss": 3.225370178222656,
"step": 44100
},
{
"epoch": 0.09655635685438177,
"grad_norm": 0.14378058910369873,
"learning_rate": 0.003,
"loss": 3.231093444824219,
"step": 44200
},
{
"epoch": 0.0967748101504324,
"grad_norm": 0.16862083971500397,
"learning_rate": 0.003,
"loss": 3.2397946166992186,
"step": 44300
},
{
"epoch": 0.09699326344648304,
"grad_norm": 0.15142101049423218,
"learning_rate": 0.003,
"loss": 3.262004699707031,
"step": 44400
},
{
"epoch": 0.09721171674253368,
"grad_norm": 0.13464325666427612,
"learning_rate": 0.003,
"loss": 3.258009338378906,
"step": 44500
},
{
"epoch": 0.09743017003858431,
"grad_norm": 0.17103958129882812,
"learning_rate": 0.003,
"loss": 3.2329156494140623,
"step": 44600
},
{
"epoch": 0.09764862333463495,
"grad_norm": 0.14914920926094055,
"learning_rate": 0.003,
"loss": 3.2394830322265626,
"step": 44700
},
{
"epoch": 0.09786707663068558,
"grad_norm": 0.12734802067279816,
"learning_rate": 0.003,
"loss": 3.2912030029296875,
"step": 44800
},
{
"epoch": 0.09808552992673622,
"grad_norm": 0.17552873492240906,
"learning_rate": 0.003,
"loss": 3.2498040771484376,
"step": 44900
},
{
"epoch": 0.09830398322278687,
"grad_norm": 0.13984735310077667,
"learning_rate": 0.003,
"loss": 3.288653564453125,
"step": 45000
},
{
"epoch": 0.0985224365188375,
"grad_norm": 0.19741426408290863,
"learning_rate": 0.003,
"loss": 3.252799377441406,
"step": 45100
},
{
"epoch": 0.09874088981488814,
"grad_norm": 0.16611187160015106,
"learning_rate": 0.003,
"loss": 3.2686611938476564,
"step": 45200
},
{
"epoch": 0.09895934311093878,
"grad_norm": 0.170828178524971,
"learning_rate": 0.003,
"loss": 3.2928143310546876,
"step": 45300
},
{
"epoch": 0.09917779640698941,
"grad_norm": 0.17449995875358582,
"learning_rate": 0.003,
"loss": 3.2658261108398436,
"step": 45400
},
{
"epoch": 0.09939624970304005,
"grad_norm": 0.19020019471645355,
"learning_rate": 0.003,
"loss": 3.285666198730469,
"step": 45500
},
{
"epoch": 0.09961470299909068,
"grad_norm": 0.14907898008823395,
"learning_rate": 0.003,
"loss": 3.2670388793945313,
"step": 45600
},
{
"epoch": 0.09983315629514132,
"grad_norm": 0.13812223076820374,
"learning_rate": 0.003,
"loss": 3.287166748046875,
"step": 45700
},
{
"epoch": 0.10005160959119196,
"grad_norm": 0.3001323342323303,
"learning_rate": 0.003,
"loss": 3.2805682373046876,
"step": 45800
},
{
"epoch": 0.1002700628872426,
"grad_norm": 0.1310071498155594,
"learning_rate": 0.003,
"loss": 3.266058044433594,
"step": 45900
},
{
"epoch": 0.10048851618329324,
"grad_norm": 0.16892723739147186,
"learning_rate": 0.003,
"loss": 3.24665283203125,
"step": 46000
},
{
"epoch": 0.10070696947934388,
"grad_norm": 0.1784011870622635,
"learning_rate": 0.003,
"loss": 3.2515863037109374,
"step": 46100
},
{
"epoch": 0.10092542277539451,
"grad_norm": 0.13655558228492737,
"learning_rate": 0.003,
"loss": 3.258023986816406,
"step": 46200
},
{
"epoch": 0.10114387607144515,
"grad_norm": 0.1459612399339676,
"learning_rate": 0.003,
"loss": 3.259796142578125,
"step": 46300
},
{
"epoch": 0.10136232936749578,
"grad_norm": 0.13782542943954468,
"learning_rate": 0.003,
"loss": 3.270685119628906,
"step": 46400
},
{
"epoch": 0.10158078266354642,
"grad_norm": 0.17929984629154205,
"learning_rate": 0.003,
"loss": 3.225979919433594,
"step": 46500
},
{
"epoch": 0.10179923595959706,
"grad_norm": 0.14070963859558105,
"learning_rate": 0.003,
"loss": 3.21829833984375,
"step": 46600
},
{
"epoch": 0.1020176892556477,
"grad_norm": 0.13950449228286743,
"learning_rate": 0.003,
"loss": 3.2376116943359374,
"step": 46700
},
{
"epoch": 0.10223614255169834,
"grad_norm": 0.14981812238693237,
"learning_rate": 0.003,
"loss": 3.2612380981445312,
"step": 46800
},
{
"epoch": 0.10245459584774898,
"grad_norm": 0.212712824344635,
"learning_rate": 0.003,
"loss": 3.221396179199219,
"step": 46900
},
{
"epoch": 0.10267304914379961,
"grad_norm": 0.1592969000339508,
"learning_rate": 0.003,
"loss": 3.334166259765625,
"step": 47000
},
{
"epoch": 0.10289150243985025,
"grad_norm": 0.14095208048820496,
"learning_rate": 0.003,
"loss": 3.24810791015625,
"step": 47100
},
{
"epoch": 0.10310995573590089,
"grad_norm": 0.36896562576293945,
"learning_rate": 0.003,
"loss": 3.1982489013671875,
"step": 47200
},
{
"epoch": 0.10332840903195152,
"grad_norm": 0.15950943529605865,
"learning_rate": 0.003,
"loss": 3.309072265625,
"step": 47300
},
{
"epoch": 0.10354686232800216,
"grad_norm": 0.15521414577960968,
"learning_rate": 0.003,
"loss": 3.28739501953125,
"step": 47400
},
{
"epoch": 0.1037653156240528,
"grad_norm": 0.16319608688354492,
"learning_rate": 0.003,
"loss": 3.276689147949219,
"step": 47500
},
{
"epoch": 0.10398376892010344,
"grad_norm": 0.18697327375411987,
"learning_rate": 0.003,
"loss": 3.27238037109375,
"step": 47600
},
{
"epoch": 0.10420222221615408,
"grad_norm": 0.1398412436246872,
"learning_rate": 0.003,
"loss": 3.27921142578125,
"step": 47700
},
{
"epoch": 0.10442067551220471,
"grad_norm": 0.14358995854854584,
"learning_rate": 0.003,
"loss": 3.2454815673828126,
"step": 47800
},
{
"epoch": 0.10463912880825535,
"grad_norm": 0.16363677382469177,
"learning_rate": 0.003,
"loss": 3.2582858276367186,
"step": 47900
},
{
"epoch": 0.10485758210430599,
"grad_norm": 0.17675544321537018,
"learning_rate": 0.003,
"loss": 3.2601687622070314,
"step": 48000
},
{
"epoch": 0.10507603540035662,
"grad_norm": 0.18376107513904572,
"learning_rate": 0.003,
"loss": 3.235495910644531,
"step": 48100
},
{
"epoch": 0.10529448869640726,
"grad_norm": 0.21416816115379333,
"learning_rate": 0.003,
"loss": 3.249112243652344,
"step": 48200
},
{
"epoch": 0.1055129419924579,
"grad_norm": 0.16644176840782166,
"learning_rate": 0.003,
"loss": 3.288675537109375,
"step": 48300
},
{
"epoch": 0.10573139528850854,
"grad_norm": 0.18745599687099457,
"learning_rate": 0.003,
"loss": 3.2637869262695314,
"step": 48400
},
{
"epoch": 0.10594984858455918,
"grad_norm": 0.1422777622938156,
"learning_rate": 0.003,
"loss": 3.301553039550781,
"step": 48500
},
{
"epoch": 0.10616830188060981,
"grad_norm": 0.18905463814735413,
"learning_rate": 0.003,
"loss": 3.28099365234375,
"step": 48600
},
{
"epoch": 0.10638675517666045,
"grad_norm": 0.16725865006446838,
"learning_rate": 0.003,
"loss": 3.289917297363281,
"step": 48700
},
{
"epoch": 0.10660520847271109,
"grad_norm": 0.12857334315776825,
"learning_rate": 0.003,
"loss": 3.2987991333007813,
"step": 48800
},
{
"epoch": 0.10682366176876172,
"grad_norm": 0.16122519969940186,
"learning_rate": 0.003,
"loss": 3.680093078613281,
"step": 48900
},
{
"epoch": 0.10704211506481236,
"grad_norm": 0.14388087391853333,
"learning_rate": 0.003,
"loss": 3.25523681640625,
"step": 49000
},
{
"epoch": 0.10726056836086299,
"grad_norm": 0.16668033599853516,
"learning_rate": 0.003,
"loss": 3.2422891235351563,
"step": 49100
},
{
"epoch": 0.10747902165691364,
"grad_norm": 0.13040824234485626,
"learning_rate": 0.003,
"loss": 3.296384582519531,
"step": 49200
},
{
"epoch": 0.10769747495296428,
"grad_norm": 0.13599610328674316,
"learning_rate": 0.003,
"loss": 3.2640185546875,
"step": 49300
},
{
"epoch": 0.10791592824901491,
"grad_norm": 0.13898426294326782,
"learning_rate": 0.003,
"loss": 3.2328057861328126,
"step": 49400
},
{
"epoch": 0.10813438154506555,
"grad_norm": 0.14450238645076752,
"learning_rate": 0.003,
"loss": 3.2923873901367187,
"step": 49500
},
{
"epoch": 0.10835283484111619,
"grad_norm": 0.14214539527893066,
"learning_rate": 0.003,
"loss": 3.273054504394531,
"step": 49600
},
{
"epoch": 0.10857128813716682,
"grad_norm": 0.15325883030891418,
"learning_rate": 0.003,
"loss": 3.2763714599609375,
"step": 49700
},
{
"epoch": 0.10878974143321746,
"grad_norm": 0.13748377561569214,
"learning_rate": 0.003,
"loss": 3.2571786499023436,
"step": 49800
},
{
"epoch": 0.10900819472926809,
"grad_norm": 0.13714741170406342,
"learning_rate": 0.003,
"loss": 3.261249694824219,
"step": 49900
},
{
"epoch": 0.10922664802531874,
"grad_norm": 0.19914014637470245,
"learning_rate": 0.003,
"loss": 3.2609954833984376,
"step": 50000
},
{
"epoch": 0.10944510132136938,
"grad_norm": 0.13507208228111267,
"learning_rate": 0.003,
"loss": 3.2500015258789063,
"step": 50100
},
{
"epoch": 0.10966355461742001,
"grad_norm": 0.15143956243991852,
"learning_rate": 0.003,
"loss": 3.23558837890625,
"step": 50200
},
{
"epoch": 0.10988200791347065,
"grad_norm": 0.13524244725704193,
"learning_rate": 0.003,
"loss": 3.29155029296875,
"step": 50300
},
{
"epoch": 0.11010046120952129,
"grad_norm": 0.13512593507766724,
"learning_rate": 0.003,
"loss": 3.2667276000976564,
"step": 50400
},
{
"epoch": 0.11031891450557192,
"grad_norm": 0.20589660108089447,
"learning_rate": 0.003,
"loss": 3.205666198730469,
"step": 50500
},
{
"epoch": 0.11053736780162256,
"grad_norm": 0.1325950175523758,
"learning_rate": 0.003,
"loss": 3.280411682128906,
"step": 50600
},
{
"epoch": 0.1107558210976732,
"grad_norm": 0.20077379047870636,
"learning_rate": 0.003,
"loss": 3.2634320068359375,
"step": 50700
},
{
"epoch": 0.11097427439372384,
"grad_norm": 0.13376857340335846,
"learning_rate": 0.003,
"loss": 3.2247994995117186,
"step": 50800
},
{
"epoch": 0.11119272768977448,
"grad_norm": 0.12860968708992004,
"learning_rate": 0.003,
"loss": 3.2442669677734375,
"step": 50900
},
{
"epoch": 0.11141118098582511,
"grad_norm": 0.20679087936878204,
"learning_rate": 0.003,
"loss": 3.266549072265625,
"step": 51000
},
{
"epoch": 0.11162963428187575,
"grad_norm": 0.15940430760383606,
"learning_rate": 0.003,
"loss": 3.2117828369140624,
"step": 51100
},
{
"epoch": 0.11184808757792639,
"grad_norm": 0.4413966238498688,
"learning_rate": 0.003,
"loss": 3.2566973876953127,
"step": 51200
},
{
"epoch": 0.11206654087397702,
"grad_norm": 0.13706789910793304,
"learning_rate": 0.003,
"loss": 3.225574645996094,
"step": 51300
},
{
"epoch": 0.11228499417002766,
"grad_norm": 0.16600839793682098,
"learning_rate": 0.003,
"loss": 3.226170654296875,
"step": 51400
},
{
"epoch": 0.1125034474660783,
"grad_norm": 0.1571975201368332,
"learning_rate": 0.003,
"loss": 3.243815612792969,
"step": 51500
},
{
"epoch": 0.11272190076212894,
"grad_norm": 0.15486016869544983,
"learning_rate": 0.003,
"loss": 3.2734530639648436,
"step": 51600
},
{
"epoch": 0.11294035405817958,
"grad_norm": 0.248945951461792,
"learning_rate": 0.003,
"loss": 3.2460952758789063,
"step": 51700
},
{
"epoch": 0.11315880735423021,
"grad_norm": 0.3302399814128876,
"learning_rate": 0.003,
"loss": 3.2613720703125,
"step": 51800
},
{
"epoch": 0.11337726065028085,
"grad_norm": 0.13467846810817719,
"learning_rate": 0.003,
"loss": 3.2903326416015624,
"step": 51900
},
{
"epoch": 0.11359571394633149,
"grad_norm": 0.16056181490421295,
"learning_rate": 0.003,
"loss": 3.26580810546875,
"step": 52000
},
{
"epoch": 0.11381416724238212,
"grad_norm": 0.246822327375412,
"learning_rate": 0.003,
"loss": 3.2113882446289064,
"step": 52100
},
{
"epoch": 0.11403262053843276,
"grad_norm": 0.147608682513237,
"learning_rate": 0.003,
"loss": 3.24245849609375,
"step": 52200
},
{
"epoch": 0.1142510738344834,
"grad_norm": 0.23047684133052826,
"learning_rate": 0.003,
"loss": 3.2903311157226565,
"step": 52300
},
{
"epoch": 0.11446952713053403,
"grad_norm": 0.2878482937812805,
"learning_rate": 0.003,
"loss": 3.236291809082031,
"step": 52400
},
{
"epoch": 0.11468798042658468,
"grad_norm": 0.2174333930015564,
"learning_rate": 0.003,
"loss": 3.265153503417969,
"step": 52500
},
{
"epoch": 0.11490643372263531,
"grad_norm": 0.12833909690380096,
"learning_rate": 0.003,
"loss": 3.2255078125,
"step": 52600
},
{
"epoch": 0.11512488701868595,
"grad_norm": 0.14357781410217285,
"learning_rate": 0.003,
"loss": 3.230517578125,
"step": 52700
},
{
"epoch": 0.11534334031473659,
"grad_norm": 0.15023696422576904,
"learning_rate": 0.003,
"loss": 3.2729867553710936,
"step": 52800
},
{
"epoch": 0.11556179361078722,
"grad_norm": 0.15742431581020355,
"learning_rate": 0.003,
"loss": 3.247342224121094,
"step": 52900
},
{
"epoch": 0.11578024690683786,
"grad_norm": 0.20336998999118805,
"learning_rate": 0.003,
"loss": 3.2814089965820314,
"step": 53000
},
{
"epoch": 0.1159987002028885,
"grad_norm": 3.120046377182007,
"learning_rate": 0.003,
"loss": 3.1381329345703124,
"step": 53100
},
{
"epoch": 0.11621715349893913,
"grad_norm": 0.15897803008556366,
"learning_rate": 0.003,
"loss": 3.2455206298828125,
"step": 53200
},
{
"epoch": 0.11643560679498978,
"grad_norm": 0.13831640779972076,
"learning_rate": 0.003,
"loss": 3.26222900390625,
"step": 53300
},
{
"epoch": 0.11665406009104042,
"grad_norm": 0.15329770743846893,
"learning_rate": 0.003,
"loss": 3.2434573364257813,
"step": 53400
},
{
"epoch": 0.11687251338709105,
"grad_norm": 0.14566101133823395,
"learning_rate": 0.003,
"loss": 3.251831359863281,
"step": 53500
},
{
"epoch": 0.11709096668314169,
"grad_norm": 0.15514326095581055,
"learning_rate": 0.003,
"loss": 3.231817321777344,
"step": 53600
},
{
"epoch": 0.11730941997919232,
"grad_norm": 0.17874790728092194,
"learning_rate": 0.003,
"loss": 3.2658575439453124,
"step": 53700
},
{
"epoch": 0.11752787327524296,
"grad_norm": 0.14967915415763855,
"learning_rate": 0.003,
"loss": 3.265608825683594,
"step": 53800
},
{
"epoch": 0.1177463265712936,
"grad_norm": 0.17889386415481567,
"learning_rate": 0.003,
"loss": 3.2492279052734374,
"step": 53900
},
{
"epoch": 0.11796477986734423,
"grad_norm": 0.18566641211509705,
"learning_rate": 0.003,
"loss": 3.2798150634765624,
"step": 54000
},
{
"epoch": 0.11818323316339488,
"grad_norm": 0.13417650759220123,
"learning_rate": 0.003,
"loss": 3.296250915527344,
"step": 54100
},
{
"epoch": 0.11840168645944552,
"grad_norm": 0.2071797102689743,
"learning_rate": 0.003,
"loss": 3.2186236572265625,
"step": 54200
},
{
"epoch": 0.11862013975549615,
"grad_norm": 0.16677580773830414,
"learning_rate": 0.003,
"loss": 3.251484680175781,
"step": 54300
},
{
"epoch": 0.11883859305154679,
"grad_norm": 0.14466318488121033,
"learning_rate": 0.003,
"loss": 3.2654324340820313,
"step": 54400
},
{
"epoch": 0.11905704634759742,
"grad_norm": 0.1548251360654831,
"learning_rate": 0.003,
"loss": 3.2652987670898437,
"step": 54500
},
{
"epoch": 0.11927549964364806,
"grad_norm": 0.15050005912780762,
"learning_rate": 0.003,
"loss": 3.251531982421875,
"step": 54600
},
{
"epoch": 0.1194939529396987,
"grad_norm": 0.15156231820583344,
"learning_rate": 0.003,
"loss": 3.2532913208007814,
"step": 54700
},
{
"epoch": 0.11971240623574933,
"grad_norm": 0.12596410512924194,
"learning_rate": 0.003,
"loss": 3.250552978515625,
"step": 54800
},
{
"epoch": 0.11993085953179998,
"grad_norm": 0.15314988791942596,
"learning_rate": 0.003,
"loss": 3.24099609375,
"step": 54900
},
{
"epoch": 0.12014931282785062,
"grad_norm": 0.14577282965183258,
"learning_rate": 0.003,
"loss": 3.2654632568359374,
"step": 55000
},
{
"epoch": 0.12036776612390125,
"grad_norm": 0.1495250016450882,
"learning_rate": 0.003,
"loss": 3.22568603515625,
"step": 55100
},
{
"epoch": 0.12058621941995189,
"grad_norm": 0.3873007595539093,
"learning_rate": 0.003,
"loss": 3.2341241455078125,
"step": 55200
},
{
"epoch": 0.12080467271600252,
"grad_norm": 0.18277539312839508,
"learning_rate": 0.003,
"loss": 3.24637939453125,
"step": 55300
},
{
"epoch": 0.12102312601205316,
"grad_norm": 0.16159602999687195,
"learning_rate": 0.003,
"loss": 3.26673828125,
"step": 55400
},
{
"epoch": 0.1212415793081038,
"grad_norm": 0.13732564449310303,
"learning_rate": 0.003,
"loss": 3.2615036010742187,
"step": 55500
},
{
"epoch": 0.12146003260415443,
"grad_norm": 0.16723576188087463,
"learning_rate": 0.003,
"loss": 3.2518795776367186,
"step": 55600
},
{
"epoch": 0.12167848590020507,
"grad_norm": 0.1542261242866516,
"learning_rate": 0.003,
"loss": 3.255987548828125,
"step": 55700
},
{
"epoch": 0.12189693919625572,
"grad_norm": 0.6683785319328308,
"learning_rate": 0.003,
"loss": 3.085527038574219,
"step": 55800
},
{
"epoch": 0.12211539249230635,
"grad_norm": 0.16766637563705444,
"learning_rate": 0.003,
"loss": 2.9290505981445314,
"step": 55900
},
{
"epoch": 0.12233384578835699,
"grad_norm": 0.15246568620204926,
"learning_rate": 0.003,
"loss": 3.2697900390625,
"step": 56000
},
{
"epoch": 0.12255229908440762,
"grad_norm": 0.16947342455387115,
"learning_rate": 0.003,
"loss": 3.2785888671875,
"step": 56100
},
{
"epoch": 0.12277075238045826,
"grad_norm": 0.14360709488391876,
"learning_rate": 0.003,
"loss": 3.2266140747070313,
"step": 56200
},
{
"epoch": 0.1229892056765089,
"grad_norm": 0.18063582479953766,
"learning_rate": 0.003,
"loss": 3.2501126098632813,
"step": 56300
},
{
"epoch": 0.12320765897255953,
"grad_norm": 0.20421698689460754,
"learning_rate": 0.003,
"loss": 3.2473678588867188,
"step": 56400
},
{
"epoch": 0.12342611226861017,
"grad_norm": 0.13375893235206604,
"learning_rate": 0.003,
"loss": 3.2326312255859375,
"step": 56500
},
{
"epoch": 0.12364456556466082,
"grad_norm": 0.14049290120601654,
"learning_rate": 0.003,
"loss": 3.2795941162109377,
"step": 56600
},
{
"epoch": 0.12386301886071145,
"grad_norm": 0.1345169097185135,
"learning_rate": 0.003,
"loss": 3.2540447998046873,
"step": 56700
},
{
"epoch": 0.12408147215676209,
"grad_norm": 0.17882542312145233,
"learning_rate": 0.003,
"loss": 3.2614974975585938,
"step": 56800
},
{
"epoch": 0.12429992545281272,
"grad_norm": 0.14673982560634613,
"learning_rate": 0.003,
"loss": 3.2368304443359377,
"step": 56900
},
{
"epoch": 0.12451837874886336,
"grad_norm": 0.15210740268230438,
"learning_rate": 0.003,
"loss": 3.2766180419921875,
"step": 57000
},
{
"epoch": 0.124736832044914,
"grad_norm": 0.2170865833759308,
"learning_rate": 0.003,
"loss": 3.2546124267578125,
"step": 57100
},
{
"epoch": 0.12495528534096463,
"grad_norm": 0.16334885358810425,
"learning_rate": 0.003,
"loss": 3.27398193359375,
"step": 57200
},
{
"epoch": 0.12517373863701528,
"grad_norm": 0.19520652294158936,
"learning_rate": 0.003,
"loss": 3.2867977905273436,
"step": 57300
},
{
"epoch": 0.12539219193306592,
"grad_norm": 0.1792980581521988,
"learning_rate": 0.003,
"loss": 3.25091552734375,
"step": 57400
},
{
"epoch": 0.12561064522911655,
"grad_norm": 0.14307785034179688,
"learning_rate": 0.003,
"loss": 3.260938720703125,
"step": 57500
},
{
"epoch": 0.1258290985251672,
"grad_norm": 0.1585034430027008,
"learning_rate": 0.003,
"loss": 3.2435162353515623,
"step": 57600
},
{
"epoch": 0.12604755182121782,
"grad_norm": 0.22285203635692596,
"learning_rate": 0.003,
"loss": 3.2605624389648438,
"step": 57700
},
{
"epoch": 0.12626600511726846,
"grad_norm": 0.16076688468456268,
"learning_rate": 0.003,
"loss": 3.247212219238281,
"step": 57800
},
{
"epoch": 0.1264844584133191,
"grad_norm": 0.21419702470302582,
"learning_rate": 0.003,
"loss": 3.258780212402344,
"step": 57900
},
{
"epoch": 0.12670291170936973,
"grad_norm": 0.129261776804924,
"learning_rate": 0.003,
"loss": 3.2596731567382813,
"step": 58000
},
{
"epoch": 0.12692136500542037,
"grad_norm": 0.13298629224300385,
"learning_rate": 0.003,
"loss": 3.2442742919921876,
"step": 58100
},
{
"epoch": 0.127139818301471,
"grad_norm": 0.1973569393157959,
"learning_rate": 0.003,
"loss": 3.2628631591796875,
"step": 58200
},
{
"epoch": 0.12735827159752164,
"grad_norm": 0.14654108881950378,
"learning_rate": 0.003,
"loss": 3.2398953247070312,
"step": 58300
},
{
"epoch": 0.12757672489357227,
"grad_norm": 0.17198579013347626,
"learning_rate": 0.003,
"loss": 3.226218566894531,
"step": 58400
},
{
"epoch": 0.1277951781896229,
"grad_norm": 0.13534477353096008,
"learning_rate": 0.003,
"loss": 3.2574343872070313,
"step": 58500
},
{
"epoch": 0.12801363148567357,
"grad_norm": 0.1584809124469757,
"learning_rate": 0.003,
"loss": 3.2984182739257815,
"step": 58600
},
{
"epoch": 0.1282320847817242,
"grad_norm": 0.173396497964859,
"learning_rate": 0.003,
"loss": 3.260642395019531,
"step": 58700
},
{
"epoch": 0.12845053807777485,
"grad_norm": 0.15514875948429108,
"learning_rate": 0.003,
"loss": 3.234716796875,
"step": 58800
},
{
"epoch": 0.12866899137382548,
"grad_norm": 0.14235849678516388,
"learning_rate": 0.003,
"loss": 3.2418527221679687,
"step": 58900
},
{
"epoch": 0.12888744466987612,
"grad_norm": 0.16027233004570007,
"learning_rate": 0.003,
"loss": 3.2690643310546874,
"step": 59000
},
{
"epoch": 0.12910589796592675,
"grad_norm": 0.14568781852722168,
"learning_rate": 0.003,
"loss": 3.2454473876953127,
"step": 59100
},
{
"epoch": 0.1293243512619774,
"grad_norm": 0.20608089864253998,
"learning_rate": 0.003,
"loss": 3.217125244140625,
"step": 59200
},
{
"epoch": 0.12954280455802802,
"grad_norm": 0.1591549515724182,
"learning_rate": 0.003,
"loss": 3.2709585571289064,
"step": 59300
},
{
"epoch": 0.12976125785407866,
"grad_norm": 0.134972482919693,
"learning_rate": 0.003,
"loss": 3.28088623046875,
"step": 59400
},
{
"epoch": 0.1299797111501293,
"grad_norm": 0.15492945909500122,
"learning_rate": 0.003,
"loss": 3.2515249633789063,
"step": 59500
},
{
"epoch": 0.13019816444617993,
"grad_norm": 0.1633273810148239,
"learning_rate": 0.003,
"loss": 3.2070086669921873,
"step": 59600
},
{
"epoch": 0.13041661774223057,
"grad_norm": 0.14716912806034088,
"learning_rate": 0.003,
"loss": 3.2317816162109376,
"step": 59700
},
{
"epoch": 0.1306350710382812,
"grad_norm": 0.16085785627365112,
"learning_rate": 0.003,
"loss": 3.25274658203125,
"step": 59800
},
{
"epoch": 0.13085352433433184,
"grad_norm": 0.16605177521705627,
"learning_rate": 0.003,
"loss": 3.263970947265625,
"step": 59900
},
{
"epoch": 0.13107197763038247,
"grad_norm": 0.16785955429077148,
"learning_rate": 0.003,
"loss": 3.22025634765625,
"step": 60000
},
{
"epoch": 0.1312904309264331,
"grad_norm": 0.1412438452243805,
"learning_rate": 0.003,
"loss": 3.220314636230469,
"step": 60100
},
{
"epoch": 0.13150888422248375,
"grad_norm": 0.21565914154052734,
"learning_rate": 0.003,
"loss": 3.229644775390625,
"step": 60200
},
{
"epoch": 0.1317273375185344,
"grad_norm": 0.19567877054214478,
"learning_rate": 0.003,
"loss": 3.249039306640625,
"step": 60300
},
{
"epoch": 0.13194579081458505,
"grad_norm": 0.17464599013328552,
"learning_rate": 0.003,
"loss": 3.2618002319335937,
"step": 60400
},
{
"epoch": 0.13216424411063568,
"grad_norm": 0.1627679020166397,
"learning_rate": 0.003,
"loss": 3.180228576660156,
"step": 60500
},
{
"epoch": 0.13238269740668632,
"grad_norm": 0.18486280739307404,
"learning_rate": 0.003,
"loss": 3.2228204345703126,
"step": 60600
},
{
"epoch": 0.13260115070273695,
"grad_norm": 0.22529979050159454,
"learning_rate": 0.003,
"loss": 3.2560736083984376,
"step": 60700
},
{
"epoch": 0.1328196039987876,
"grad_norm": 0.19143950939178467,
"learning_rate": 0.003,
"loss": 3.2156741333007814,
"step": 60800
},
{
"epoch": 0.13303805729483822,
"grad_norm": 0.1540822684764862,
"learning_rate": 0.003,
"loss": 3.1913638305664063,
"step": 60900
},
{
"epoch": 0.13325651059088886,
"grad_norm": 0.13871780037879944,
"learning_rate": 0.003,
"loss": 3.2044723510742186,
"step": 61000
},
{
"epoch": 0.1334749638869395,
"grad_norm": 0.14368896186351776,
"learning_rate": 0.003,
"loss": 3.2619207763671874,
"step": 61100
},
{
"epoch": 0.13369341718299013,
"grad_norm": 0.16204333305358887,
"learning_rate": 0.003,
"loss": 3.2338299560546875,
"step": 61200
},
{
"epoch": 0.13391187047904077,
"grad_norm": 0.13743866980075836,
"learning_rate": 0.003,
"loss": 3.2367453002929687,
"step": 61300
},
{
"epoch": 0.1341303237750914,
"grad_norm": 0.14885108172893524,
"learning_rate": 0.003,
"loss": 3.2349374389648435,
"step": 61400
},
{
"epoch": 0.13434877707114204,
"grad_norm": 0.17403972148895264,
"learning_rate": 0.003,
"loss": 3.24302734375,
"step": 61500
},
{
"epoch": 0.13456723036719268,
"grad_norm": 0.15033158659934998,
"learning_rate": 0.003,
"loss": 3.263538818359375,
"step": 61600
},
{
"epoch": 0.1347856836632433,
"grad_norm": 0.21722860634326935,
"learning_rate": 0.003,
"loss": 3.2349429321289063,
"step": 61700
},
{
"epoch": 0.13500413695929395,
"grad_norm": 0.1466379314661026,
"learning_rate": 0.003,
"loss": 3.219266052246094,
"step": 61800
},
{
"epoch": 0.1352225902553446,
"grad_norm": 0.13940411806106567,
"learning_rate": 0.003,
"loss": 3.2392041015625,
"step": 61900
},
{
"epoch": 0.13544104355139525,
"grad_norm": 0.13754943013191223,
"learning_rate": 0.003,
"loss": 3.252232971191406,
"step": 62000
},
{
"epoch": 0.13565949684744588,
"grad_norm": 0.32490044832229614,
"learning_rate": 0.003,
"loss": 3.234407653808594,
"step": 62100
},
{
"epoch": 0.13587795014349652,
"grad_norm": 0.14267276227474213,
"learning_rate": 0.003,
"loss": 3.218851318359375,
"step": 62200
},
{
"epoch": 0.13609640343954715,
"grad_norm": 0.21616226434707642,
"learning_rate": 0.003,
"loss": 3.2713690185546875,
"step": 62300
},
{
"epoch": 0.1363148567355978,
"grad_norm": 0.1464727371931076,
"learning_rate": 0.003,
"loss": 3.24201416015625,
"step": 62400
},
{
"epoch": 0.13653331003164842,
"grad_norm": 0.5176395177841187,
"learning_rate": 0.003,
"loss": 3.2492108154296875,
"step": 62500
},
{
"epoch": 0.13675176332769906,
"grad_norm": 0.1350073516368866,
"learning_rate": 0.003,
"loss": 3.2798583984375,
"step": 62600
},
{
"epoch": 0.1369702166237497,
"grad_norm": 0.18025344610214233,
"learning_rate": 0.003,
"loss": 3.2309701538085935,
"step": 62700
},
{
"epoch": 0.13718866991980033,
"grad_norm": 0.18821361660957336,
"learning_rate": 0.003,
"loss": 3.2689004516601563,
"step": 62800
},
{
"epoch": 0.13740712321585097,
"grad_norm": 0.1663011610507965,
"learning_rate": 0.003,
"loss": 3.2441537475585935,
"step": 62900
},
{
"epoch": 0.1376255765119016,
"grad_norm": 0.15948191285133362,
"learning_rate": 0.003,
"loss": 2.7289602661132815,
"step": 63000
},
{
"epoch": 0.13784402980795224,
"grad_norm": 0.1529848873615265,
"learning_rate": 0.003,
"loss": 3.2516909790039064,
"step": 63100
},
{
"epoch": 0.13806248310400288,
"grad_norm": 0.14751732349395752,
"learning_rate": 0.003,
"loss": 3.246956481933594,
"step": 63200
},
{
"epoch": 0.1382809364000535,
"grad_norm": 0.1541360467672348,
"learning_rate": 0.003,
"loss": 3.2233453369140626,
"step": 63300
},
{
"epoch": 0.13849938969610415,
"grad_norm": 0.17564933001995087,
"learning_rate": 0.003,
"loss": 3.2218228149414063,
"step": 63400
},
{
"epoch": 0.13871784299215478,
"grad_norm": 0.14369294047355652,
"learning_rate": 0.003,
"loss": 3.3964053344726564,
"step": 63500
},
{
"epoch": 0.13893629628820545,
"grad_norm": 0.14323627948760986,
"learning_rate": 0.003,
"loss": 3.249570617675781,
"step": 63600
},
{
"epoch": 0.13915474958425608,
"grad_norm": 0.1641722321510315,
"learning_rate": 0.003,
"loss": 3.1940841674804688,
"step": 63700
},
{
"epoch": 0.13937320288030672,
"grad_norm": 0.15655729174613953,
"learning_rate": 0.003,
"loss": 3.230262451171875,
"step": 63800
},
{
"epoch": 0.13959165617635735,
"grad_norm": 0.23219190537929535,
"learning_rate": 0.003,
"loss": 3.260406494140625,
"step": 63900
},
{
"epoch": 0.139810109472408,
"grad_norm": 0.15642347931861877,
"learning_rate": 0.003,
"loss": 3.266820068359375,
"step": 64000
},
{
"epoch": 0.14002856276845863,
"grad_norm": 0.17600469291210175,
"learning_rate": 0.003,
"loss": 3.214679870605469,
"step": 64100
},
{
"epoch": 0.14024701606450926,
"grad_norm": 0.1333693265914917,
"learning_rate": 0.003,
"loss": 3.217235107421875,
"step": 64200
},
{
"epoch": 0.1404654693605599,
"grad_norm": 0.17207157611846924,
"learning_rate": 0.003,
"loss": 3.2215121459960936,
"step": 64300
},
{
"epoch": 0.14068392265661053,
"grad_norm": 0.14259660243988037,
"learning_rate": 0.003,
"loss": 3.2310165405273437,
"step": 64400
},
{
"epoch": 0.14090237595266117,
"grad_norm": 0.16601745784282684,
"learning_rate": 0.003,
"loss": 3.2018670654296875,
"step": 64500
},
{
"epoch": 0.1411208292487118,
"grad_norm": 0.14996014535427094,
"learning_rate": 0.003,
"loss": 3.21740234375,
"step": 64600
},
{
"epoch": 0.14133928254476244,
"grad_norm": 0.16059845685958862,
"learning_rate": 0.003,
"loss": 3.204219970703125,
"step": 64700
},
{
"epoch": 0.14155773584081308,
"grad_norm": 0.18016761541366577,
"learning_rate": 0.003,
"loss": 3.195794982910156,
"step": 64800
},
{
"epoch": 0.1417761891368637,
"grad_norm": 0.15482042729854584,
"learning_rate": 0.003,
"loss": 3.209017333984375,
"step": 64900
},
{
"epoch": 0.14199464243291435,
"grad_norm": 0.19144226610660553,
"learning_rate": 0.003,
"loss": 3.210953674316406,
"step": 65000
},
{
"epoch": 0.14221309572896498,
"grad_norm": 0.14203697443008423,
"learning_rate": 0.003,
"loss": 3.19835205078125,
"step": 65100
},
{
"epoch": 0.14243154902501565,
"grad_norm": 0.4821944236755371,
"learning_rate": 0.003,
"loss": 3.25100341796875,
"step": 65200
},
{
"epoch": 0.14265000232106628,
"grad_norm": 0.15543989837169647,
"learning_rate": 0.003,
"loss": 3.20369873046875,
"step": 65300
},
{
"epoch": 0.14286845561711692,
"grad_norm": 0.1701357513666153,
"learning_rate": 0.003,
"loss": 3.169871520996094,
"step": 65400
},
{
"epoch": 0.14308690891316755,
"grad_norm": 0.17308828234672546,
"learning_rate": 0.003,
"loss": 3.2284805297851564,
"step": 65500
},
{
"epoch": 0.1433053622092182,
"grad_norm": 0.17479535937309265,
"learning_rate": 0.003,
"loss": 3.20006103515625,
"step": 65600
},
{
"epoch": 0.14352381550526883,
"grad_norm": 0.14500826597213745,
"learning_rate": 0.003,
"loss": 3.2878839111328126,
"step": 65700
},
{
"epoch": 0.14374226880131946,
"grad_norm": 0.1360345333814621,
"learning_rate": 0.003,
"loss": 3.29202880859375,
"step": 65800
},
{
"epoch": 0.1439607220973701,
"grad_norm": 0.21259358525276184,
"learning_rate": 0.003,
"loss": 3.2273773193359374,
"step": 65900
},
{
"epoch": 0.14417917539342073,
"grad_norm": 0.2542419731616974,
"learning_rate": 0.003,
"loss": 3.2165481567382814,
"step": 66000
},
{
"epoch": 0.14439762868947137,
"grad_norm": 0.24819156527519226,
"learning_rate": 0.003,
"loss": 3.243924560546875,
"step": 66100
},
{
"epoch": 0.144616081985522,
"grad_norm": 0.13760997354984283,
"learning_rate": 0.003,
"loss": 3.2319210815429686,
"step": 66200
},
{
"epoch": 0.14483453528157264,
"grad_norm": 0.15245407819747925,
"learning_rate": 0.003,
"loss": 3.2334552001953125,
"step": 66300
},
{
"epoch": 0.14505298857762328,
"grad_norm": 0.15227407217025757,
"learning_rate": 0.003,
"loss": 3.217420959472656,
"step": 66400
},
{
"epoch": 0.1452714418736739,
"grad_norm": 0.14392466843128204,
"learning_rate": 0.003,
"loss": 3.2076754760742188,
"step": 66500
},
{
"epoch": 0.14548989516972455,
"grad_norm": 0.21550512313842773,
"learning_rate": 0.003,
"loss": 3.266759338378906,
"step": 66600
},
{
"epoch": 0.14570834846577518,
"grad_norm": 0.16961754858493805,
"learning_rate": 0.003,
"loss": 3.2540386962890624,
"step": 66700
},
{
"epoch": 0.14592680176182582,
"grad_norm": 0.1846882551908493,
"learning_rate": 0.003,
"loss": 3.2260113525390626,
"step": 66800
},
{
"epoch": 0.14614525505787648,
"grad_norm": 0.17954540252685547,
"learning_rate": 0.003,
"loss": 3.2294110107421874,
"step": 66900
},
{
"epoch": 0.14636370835392712,
"grad_norm": 0.1706397533416748,
"learning_rate": 0.003,
"loss": 3.248121337890625,
"step": 67000
},
{
"epoch": 0.14658216164997775,
"grad_norm": 0.17211204767227173,
"learning_rate": 0.003,
"loss": 3.2395590209960936,
"step": 67100
},
{
"epoch": 0.1468006149460284,
"grad_norm": 0.21838849782943726,
"learning_rate": 0.003,
"loss": 3.241527099609375,
"step": 67200
},
{
"epoch": 0.14701906824207903,
"grad_norm": 0.19266127049922943,
"learning_rate": 0.003,
"loss": 3.2631201171875,
"step": 67300
},
{
"epoch": 0.14723752153812966,
"grad_norm": 0.15134894847869873,
"learning_rate": 0.003,
"loss": 3.2437713623046873,
"step": 67400
},
{
"epoch": 0.1474559748341803,
"grad_norm": 0.34054991602897644,
"learning_rate": 0.003,
"loss": 3.2246142578125,
"step": 67500
},
{
"epoch": 0.14767442813023093,
"grad_norm": 0.15797755122184753,
"learning_rate": 0.003,
"loss": 3.2434561157226565,
"step": 67600
},
{
"epoch": 0.14789288142628157,
"grad_norm": 0.1528821587562561,
"learning_rate": 0.003,
"loss": 3.2275845336914064,
"step": 67700
},
{
"epoch": 0.1481113347223322,
"grad_norm": 0.1330331712961197,
"learning_rate": 0.003,
"loss": 3.2324395751953126,
"step": 67800
},
{
"epoch": 0.14832978801838284,
"grad_norm": 0.19141782820224762,
"learning_rate": 0.003,
"loss": 3.2720477294921877,
"step": 67900
},
{
"epoch": 0.14854824131443348,
"grad_norm": 0.13883931934833527,
"learning_rate": 0.003,
"loss": 3.272904357910156,
"step": 68000
},
{
"epoch": 0.1487666946104841,
"grad_norm": 0.14308680593967438,
"learning_rate": 0.003,
"loss": 3.227841491699219,
"step": 68100
},
{
"epoch": 0.14898514790653475,
"grad_norm": 0.17235063016414642,
"learning_rate": 0.003,
"loss": 3.242052001953125,
"step": 68200
},
{
"epoch": 0.14920360120258538,
"grad_norm": 0.17289380729198456,
"learning_rate": 0.003,
"loss": 3.2746286010742187,
"step": 68300
},
{
"epoch": 0.14942205449863602,
"grad_norm": 0.14802849292755127,
"learning_rate": 0.003,
"loss": 3.194547424316406,
"step": 68400
},
{
"epoch": 0.14964050779468668,
"grad_norm": 0.19775991141796112,
"learning_rate": 0.003,
"loss": 3.181102600097656,
"step": 68500
},
{
"epoch": 0.14985896109073732,
"grad_norm": 0.13130341470241547,
"learning_rate": 0.003,
"loss": 2.723526611328125,
"step": 68600
},
{
"epoch": 0.15007741438678796,
"grad_norm": 0.14095982909202576,
"learning_rate": 0.003,
"loss": 3.231905212402344,
"step": 68700
},
{
"epoch": 0.1502958676828386,
"grad_norm": 0.14556872844696045,
"learning_rate": 0.003,
"loss": 3.2342495727539062,
"step": 68800
},
{
"epoch": 0.15051432097888923,
"grad_norm": 0.14050233364105225,
"learning_rate": 0.003,
"loss": 3.2132330322265625,
"step": 68900
},
{
"epoch": 0.15073277427493986,
"grad_norm": 0.1730201691389084,
"learning_rate": 0.003,
"loss": 3.2115792846679687,
"step": 69000
},
{
"epoch": 0.1509512275709905,
"grad_norm": 0.20617027580738068,
"learning_rate": 0.003,
"loss": 3.2500860595703127,
"step": 69100
},
{
"epoch": 0.15116968086704113,
"grad_norm": 0.1822213977575302,
"learning_rate": 0.003,
"loss": 3.236435852050781,
"step": 69200
},
{
"epoch": 0.15138813416309177,
"grad_norm": 0.17483313381671906,
"learning_rate": 0.003,
"loss": 3.253581237792969,
"step": 69300
},
{
"epoch": 0.1516065874591424,
"grad_norm": 0.15619760751724243,
"learning_rate": 0.003,
"loss": 3.196072692871094,
"step": 69400
},
{
"epoch": 0.15182504075519304,
"grad_norm": 0.18323157727718353,
"learning_rate": 0.003,
"loss": 3.2476675415039065,
"step": 69500
},
{
"epoch": 0.15204349405124368,
"grad_norm": 0.1582677662372589,
"learning_rate": 0.003,
"loss": 3.2430215454101563,
"step": 69600
},
{
"epoch": 0.1522619473472943,
"grad_norm": 0.15574927628040314,
"learning_rate": 0.003,
"loss": 3.219065856933594,
"step": 69700
},
{
"epoch": 0.15248040064334495,
"grad_norm": 0.13565471768379211,
"learning_rate": 0.003,
"loss": 3.2654583740234373,
"step": 69800
},
{
"epoch": 0.15269885393939558,
"grad_norm": 0.1486131101846695,
"learning_rate": 0.003,
"loss": 3.2257269287109374,
"step": 69900
},
{
"epoch": 0.15291730723544622,
"grad_norm": 0.16414867341518402,
"learning_rate": 0.003,
"loss": 3.227646484375,
"step": 70000
},
{
"epoch": 0.15313576053149686,
"grad_norm": 0.17787083983421326,
"learning_rate": 0.003,
"loss": 3.1835894775390625,
"step": 70100
},
{
"epoch": 0.15335421382754752,
"grad_norm": 0.14870117604732513,
"learning_rate": 0.003,
"loss": 3.2247634887695313,
"step": 70200
},
{
"epoch": 0.15357266712359816,
"grad_norm": 0.14349189400672913,
"learning_rate": 0.003,
"loss": 3.171583251953125,
"step": 70300
},
{
"epoch": 0.1537911204196488,
"grad_norm": 0.18492130935192108,
"learning_rate": 0.003,
"loss": 3.2198065185546874,
"step": 70400
},
{
"epoch": 0.15400957371569943,
"grad_norm": 0.18474198877811432,
"learning_rate": 0.003,
"loss": 3.2141107177734374,
"step": 70500
},
{
"epoch": 0.15422802701175006,
"grad_norm": 0.20480042695999146,
"learning_rate": 0.003,
"loss": 3.2477322387695313,
"step": 70600
},
{
"epoch": 0.1544464803078007,
"grad_norm": 0.15379810333251953,
"learning_rate": 0.003,
"loss": 3.2210733032226564,
"step": 70700
},
{
"epoch": 0.15466493360385133,
"grad_norm": 0.1564520001411438,
"learning_rate": 0.003,
"loss": 3.2100540161132813,
"step": 70800
},
{
"epoch": 0.15488338689990197,
"grad_norm": 0.1851634830236435,
"learning_rate": 0.003,
"loss": 3.2196975708007813,
"step": 70900
},
{
"epoch": 0.1551018401959526,
"grad_norm": 0.13629335165023804,
"learning_rate": 0.003,
"loss": 3.2184857177734374,
"step": 71000
},
{
"epoch": 0.15532029349200324,
"grad_norm": 0.13462699949741364,
"learning_rate": 0.003,
"loss": 3.2634429931640625,
"step": 71100
},
{
"epoch": 0.15553874678805388,
"grad_norm": 0.14812074601650238,
"learning_rate": 0.003,
"loss": 3.2562606811523436,
"step": 71200
},
{
"epoch": 0.1557572000841045,
"grad_norm": 0.13378751277923584,
"learning_rate": 0.003,
"loss": 3.2482659912109373,
"step": 71300
},
{
"epoch": 0.15597565338015515,
"grad_norm": 0.16416287422180176,
"learning_rate": 0.003,
"loss": 3.1697012329101564,
"step": 71400
},
{
"epoch": 0.15619410667620579,
"grad_norm": 0.1352565735578537,
"learning_rate": 0.003,
"loss": 3.2339556884765623,
"step": 71500
},
{
"epoch": 0.15641255997225642,
"grad_norm": 0.14081928133964539,
"learning_rate": 0.003,
"loss": 3.2458950805664064,
"step": 71600
},
{
"epoch": 0.15663101326830706,
"grad_norm": 0.1376163810491562,
"learning_rate": 0.003,
"loss": 3.21879150390625,
"step": 71700
},
{
"epoch": 0.15684946656435772,
"grad_norm": 0.16404809057712555,
"learning_rate": 0.003,
"loss": 3.261857604980469,
"step": 71800
},
{
"epoch": 0.15706791986040836,
"grad_norm": 0.16098147630691528,
"learning_rate": 0.003,
"loss": 3.2662176513671874,
"step": 71900
},
{
"epoch": 0.157286373156459,
"grad_norm": 0.19868750870227814,
"learning_rate": 0.003,
"loss": 3.25436767578125,
"step": 72000
},
{
"epoch": 0.15750482645250963,
"grad_norm": 0.18838000297546387,
"learning_rate": 0.003,
"loss": 3.2456134033203123,
"step": 72100
},
{
"epoch": 0.15772327974856026,
"grad_norm": 0.18162785470485687,
"learning_rate": 0.003,
"loss": 3.2529983520507812,
"step": 72200
},
{
"epoch": 0.1579417330446109,
"grad_norm": 0.13635674118995667,
"learning_rate": 0.003,
"loss": 3.2445623779296877,
"step": 72300
},
{
"epoch": 0.15816018634066153,
"grad_norm": 0.18136821687221527,
"learning_rate": 0.003,
"loss": 3.1944915771484377,
"step": 72400
},
{
"epoch": 0.15837863963671217,
"grad_norm": 0.16928516328334808,
"learning_rate": 0.003,
"loss": 3.2309170532226563,
"step": 72500
},
{
"epoch": 0.1585970929327628,
"grad_norm": 0.14556556940078735,
"learning_rate": 0.003,
"loss": 3.2353945922851564,
"step": 72600
},
{
"epoch": 0.15881554622881344,
"grad_norm": 0.14194248616695404,
"learning_rate": 0.003,
"loss": 3.262231750488281,
"step": 72700
},
{
"epoch": 0.15903399952486408,
"grad_norm": 0.15043506026268005,
"learning_rate": 0.003,
"loss": 3.2179226684570312,
"step": 72800
},
{
"epoch": 0.15925245282091471,
"grad_norm": 0.18483737111091614,
"learning_rate": 0.003,
"loss": 3.2244232177734373,
"step": 72900
},
{
"epoch": 0.15947090611696535,
"grad_norm": 0.17586249113082886,
"learning_rate": 0.003,
"loss": 3.1289639282226562,
"step": 73000
},
{
"epoch": 0.15968935941301599,
"grad_norm": 0.17315636575222015,
"learning_rate": 0.003,
"loss": 3.230426940917969,
"step": 73100
},
{
"epoch": 0.15990781270906662,
"grad_norm": 0.14286355674266815,
"learning_rate": 0.003,
"loss": 3.2417947387695314,
"step": 73200
},
{
"epoch": 0.16012626600511726,
"grad_norm": 0.15196926891803741,
"learning_rate": 0.003,
"loss": 3.2354244995117187,
"step": 73300
},
{
"epoch": 0.1603447193011679,
"grad_norm": 0.1744149625301361,
"learning_rate": 0.003,
"loss": 3.165083923339844,
"step": 73400
},
{
"epoch": 0.16056317259721856,
"grad_norm": 0.15688399970531464,
"learning_rate": 0.003,
"loss": 3.21979736328125,
"step": 73500
},
{
"epoch": 0.1607816258932692,
"grad_norm": 0.1604417860507965,
"learning_rate": 0.003,
"loss": 3.223596496582031,
"step": 73600
},
{
"epoch": 0.16100007918931983,
"grad_norm": 0.14714471995830536,
"learning_rate": 0.003,
"loss": 3.2240216064453127,
"step": 73700
},
{
"epoch": 0.16121853248537046,
"grad_norm": 0.18966136872768402,
"learning_rate": 0.003,
"loss": 3.197341003417969,
"step": 73800
},
{
"epoch": 0.1614369857814211,
"grad_norm": 0.2152947634458542,
"learning_rate": 0.003,
"loss": 3.1614569091796874,
"step": 73900
},
{
"epoch": 0.16165543907747174,
"grad_norm": 0.14745226502418518,
"learning_rate": 0.003,
"loss": 3.2265939331054687,
"step": 74000
},
{
"epoch": 0.16187389237352237,
"grad_norm": 0.15095406770706177,
"learning_rate": 0.003,
"loss": 3.373876037597656,
"step": 74100
},
{
"epoch": 0.162092345669573,
"grad_norm": 0.1664532572031021,
"learning_rate": 0.003,
"loss": 3.1829833984375,
"step": 74200
},
{
"epoch": 0.16231079896562364,
"grad_norm": 0.14805693924427032,
"learning_rate": 0.003,
"loss": 3.207042236328125,
"step": 74300
},
{
"epoch": 0.16252925226167428,
"grad_norm": 0.15347807109355927,
"learning_rate": 0.003,
"loss": 3.1845889282226563,
"step": 74400
},
{
"epoch": 0.16274770555772491,
"grad_norm": 0.20546002686023712,
"learning_rate": 0.003,
"loss": 3.1983770751953124,
"step": 74500
},
{
"epoch": 0.16296615885377555,
"grad_norm": 0.14666610956192017,
"learning_rate": 0.003,
"loss": 3.222270202636719,
"step": 74600
},
{
"epoch": 0.16318461214982619,
"grad_norm": 0.15613286197185516,
"learning_rate": 0.003,
"loss": 3.2073956298828126,
"step": 74700
},
{
"epoch": 0.16340306544587682,
"grad_norm": 0.16200363636016846,
"learning_rate": 0.003,
"loss": 3.2018893432617186,
"step": 74800
},
{
"epoch": 0.16362151874192746,
"grad_norm": 0.1372697651386261,
"learning_rate": 0.003,
"loss": 3.2019302368164064,
"step": 74900
},
{
"epoch": 0.1638399720379781,
"grad_norm": 0.1444454938173294,
"learning_rate": 0.003,
"loss": 3.189869079589844,
"step": 75000
},
{
"epoch": 0.16405842533402876,
"grad_norm": 0.16384346783161163,
"learning_rate": 0.003,
"loss": 3.242828674316406,
"step": 75100
},
{
"epoch": 0.1642768786300794,
"grad_norm": 0.22108645737171173,
"learning_rate": 0.003,
"loss": 3.203451843261719,
"step": 75200
},
{
"epoch": 0.16449533192613003,
"grad_norm": 0.155759796500206,
"learning_rate": 0.003,
"loss": 3.1484210205078127,
"step": 75300
},
{
"epoch": 0.16471378522218066,
"grad_norm": 0.1365097612142563,
"learning_rate": 0.003,
"loss": 3.177284851074219,
"step": 75400
},
{
"epoch": 0.1649322385182313,
"grad_norm": 0.20308803021907806,
"learning_rate": 0.003,
"loss": 3.2935577392578126,
"step": 75500
},
{
"epoch": 0.16515069181428194,
"grad_norm": 0.1627846658229828,
"learning_rate": 0.003,
"loss": 3.2038775634765626,
"step": 75600
},
{
"epoch": 0.16536914511033257,
"grad_norm": 0.15723568201065063,
"learning_rate": 0.003,
"loss": 3.18799072265625,
"step": 75700
},
{
"epoch": 0.1655875984063832,
"grad_norm": 0.14858828485012054,
"learning_rate": 0.003,
"loss": 3.2345623779296875,
"step": 75800
},
{
"epoch": 0.16580605170243384,
"grad_norm": 0.16737258434295654,
"learning_rate": 0.003,
"loss": 3.1938626098632814,
"step": 75900
},
{
"epoch": 0.16602450499848448,
"grad_norm": 0.15170836448669434,
"learning_rate": 0.003,
"loss": 3.2006201171875,
"step": 76000
},
{
"epoch": 0.16624295829453511,
"grad_norm": 0.1959700584411621,
"learning_rate": 0.003,
"loss": 3.254922180175781,
"step": 76100
},
{
"epoch": 0.16646141159058575,
"grad_norm": 0.15581128001213074,
"learning_rate": 0.003,
"loss": 3.1887762451171877,
"step": 76200
},
{
"epoch": 0.1666798648866364,
"grad_norm": 0.14930440485477448,
"learning_rate": 0.003,
"loss": 3.2389352416992185,
"step": 76300
},
{
"epoch": 0.16689831818268702,
"grad_norm": 0.2020094394683838,
"learning_rate": 0.003,
"loss": 3.2333514404296877,
"step": 76400
},
{
"epoch": 0.16711677147873766,
"grad_norm": 0.1374303102493286,
"learning_rate": 0.003,
"loss": 3.2222906494140626,
"step": 76500
},
{
"epoch": 0.1673352247747883,
"grad_norm": 0.15284472703933716,
"learning_rate": 0.003,
"loss": 3.2347198486328126,
"step": 76600
},
{
"epoch": 0.16755367807083893,
"grad_norm": 0.23065847158432007,
"learning_rate": 0.003,
"loss": 3.2416390991210937,
"step": 76700
},
{
"epoch": 0.1677721313668896,
"grad_norm": 0.15649299323558807,
"learning_rate": 0.003,
"loss": 3.2050161743164063,
"step": 76800
},
{
"epoch": 0.16799058466294023,
"grad_norm": 0.14923731982707977,
"learning_rate": 0.003,
"loss": 3.2096380615234374,
"step": 76900
},
{
"epoch": 0.16820903795899086,
"grad_norm": 0.15309712290763855,
"learning_rate": 0.003,
"loss": 3.327672424316406,
"step": 77000
},
{
"epoch": 0.1684274912550415,
"grad_norm": 0.1624584048986435,
"learning_rate": 0.003,
"loss": 3.0921026611328126,
"step": 77100
},
{
"epoch": 0.16864594455109214,
"grad_norm": 0.15112026035785675,
"learning_rate": 0.003,
"loss": 3.2112442016601563,
"step": 77200
},
{
"epoch": 0.16886439784714277,
"grad_norm": 0.2171279788017273,
"learning_rate": 0.003,
"loss": 3.2253741455078124,
"step": 77300
},
{
"epoch": 0.1690828511431934,
"grad_norm": 0.14117172360420227,
"learning_rate": 0.003,
"loss": 3.26427001953125,
"step": 77400
},
{
"epoch": 0.16930130443924404,
"grad_norm": 0.20454275608062744,
"learning_rate": 0.003,
"loss": 3.2036868286132814,
"step": 77500
},
{
"epoch": 0.16951975773529468,
"grad_norm": 0.15810075402259827,
"learning_rate": 0.003,
"loss": 3.2048870849609377,
"step": 77600
},
{
"epoch": 0.16973821103134532,
"grad_norm": 0.14809603989124298,
"learning_rate": 0.003,
"loss": 3.239774475097656,
"step": 77700
},
{
"epoch": 0.16995666432739595,
"grad_norm": 0.13929158449172974,
"learning_rate": 0.003,
"loss": 3.182627868652344,
"step": 77800
},
{
"epoch": 0.1701751176234466,
"grad_norm": 0.14580564200878143,
"learning_rate": 0.003,
"loss": 3.209283752441406,
"step": 77900
},
{
"epoch": 0.17039357091949722,
"grad_norm": 0.169144406914711,
"learning_rate": 0.003,
"loss": 3.2105447387695314,
"step": 78000
},
{
"epoch": 0.17061202421554786,
"grad_norm": 0.18892961740493774,
"learning_rate": 0.003,
"loss": 3.179049377441406,
"step": 78100
},
{
"epoch": 0.1708304775115985,
"grad_norm": 0.17873692512512207,
"learning_rate": 0.003,
"loss": 3.1997015380859377,
"step": 78200
},
{
"epoch": 0.17104893080764913,
"grad_norm": 0.16759565472602844,
"learning_rate": 0.003,
"loss": 3.2110018920898438,
"step": 78300
},
{
"epoch": 0.1712673841036998,
"grad_norm": 0.1694764345884323,
"learning_rate": 0.003,
"loss": 3.1810675048828125,
"step": 78400
},
{
"epoch": 0.17148583739975043,
"grad_norm": 0.2141251564025879,
"learning_rate": 0.003,
"loss": 3.2247308349609374,
"step": 78500
},
{
"epoch": 0.17170429069580107,
"grad_norm": 0.1924314647912979,
"learning_rate": 0.003,
"loss": 3.246143493652344,
"step": 78600
},
{
"epoch": 0.1719227439918517,
"grad_norm": 0.14677858352661133,
"learning_rate": 0.003,
"loss": 3.195514831542969,
"step": 78700
},
{
"epoch": 0.17214119728790234,
"grad_norm": 0.13610516488552094,
"learning_rate": 0.003,
"loss": 3.2029443359375,
"step": 78800
},
{
"epoch": 0.17235965058395297,
"grad_norm": 0.1633630096912384,
"learning_rate": 0.003,
"loss": 3.1546783447265625,
"step": 78900
},
{
"epoch": 0.1725781038800036,
"grad_norm": 0.1665549874305725,
"learning_rate": 0.003,
"loss": 3.2146710205078124,
"step": 79000
},
{
"epoch": 0.17279655717605424,
"grad_norm": 0.39296653866767883,
"learning_rate": 0.003,
"loss": 3.2075177001953126,
"step": 79100
},
{
"epoch": 0.17301501047210488,
"grad_norm": 0.28278303146362305,
"learning_rate": 0.003,
"loss": 3.201310729980469,
"step": 79200
},
{
"epoch": 0.17323346376815552,
"grad_norm": 0.1643471121788025,
"learning_rate": 0.003,
"loss": 3.235052490234375,
"step": 79300
},
{
"epoch": 0.17345191706420615,
"grad_norm": 0.18586060404777527,
"learning_rate": 0.003,
"loss": 3.168081970214844,
"step": 79400
},
{
"epoch": 0.1736703703602568,
"grad_norm": 0.19346100091934204,
"learning_rate": 0.003,
"loss": 3.218128662109375,
"step": 79500
},
{
"epoch": 0.17388882365630742,
"grad_norm": 0.1757037341594696,
"learning_rate": 0.003,
"loss": 3.1765069580078125,
"step": 79600
},
{
"epoch": 0.17410727695235806,
"grad_norm": 0.15886549651622772,
"learning_rate": 0.003,
"loss": 3.2562713623046875,
"step": 79700
},
{
"epoch": 0.1743257302484087,
"grad_norm": 0.16389308869838715,
"learning_rate": 0.003,
"loss": 3.220198974609375,
"step": 79800
},
{
"epoch": 0.17454418354445933,
"grad_norm": 0.30649611353874207,
"learning_rate": 0.003,
"loss": 3.2026336669921873,
"step": 79900
},
{
"epoch": 0.17476263684050997,
"grad_norm": 0.18294557929039001,
"learning_rate": 0.003,
"loss": 3.1905865478515625,
"step": 80000
},
{
"epoch": 0.17498109013656063,
"grad_norm": 0.1466442495584488,
"learning_rate": 0.003,
"loss": 3.2082867431640625,
"step": 80100
},
{
"epoch": 0.17519954343261127,
"grad_norm": 0.15893469750881195,
"learning_rate": 0.003,
"loss": 3.0881536865234374,
"step": 80200
},
{
"epoch": 0.1754179967286619,
"grad_norm": 0.15483619272708893,
"learning_rate": 0.003,
"loss": 3.192733154296875,
"step": 80300
},
{
"epoch": 0.17563645002471254,
"grad_norm": 0.19978880882263184,
"learning_rate": 0.003,
"loss": 3.2413204956054686,
"step": 80400
},
{
"epoch": 0.17585490332076317,
"grad_norm": 0.1491372138261795,
"learning_rate": 0.003,
"loss": 3.2084686279296877,
"step": 80500
},
{
"epoch": 0.1760733566168138,
"grad_norm": 0.2566303610801697,
"learning_rate": 0.003,
"loss": 3.2092715454101564,
"step": 80600
},
{
"epoch": 0.17629180991286444,
"grad_norm": 0.2119085043668747,
"learning_rate": 0.003,
"loss": 3.222015380859375,
"step": 80700
},
{
"epoch": 0.17651026320891508,
"grad_norm": 0.14177367091178894,
"learning_rate": 0.003,
"loss": 3.1890496826171875,
"step": 80800
},
{
"epoch": 0.17672871650496572,
"grad_norm": 0.1567944884300232,
"learning_rate": 0.003,
"loss": 3.248140563964844,
"step": 80900
},
{
"epoch": 0.17694716980101635,
"grad_norm": 0.13174422085285187,
"learning_rate": 0.003,
"loss": 3.19893798828125,
"step": 81000
},
{
"epoch": 0.177165623097067,
"grad_norm": 0.1621484011411667,
"learning_rate": 0.003,
"loss": 3.17727783203125,
"step": 81100
},
{
"epoch": 0.17738407639311762,
"grad_norm": 0.21841928362846375,
"learning_rate": 0.003,
"loss": 3.301322326660156,
"step": 81200
},
{
"epoch": 0.17760252968916826,
"grad_norm": 1.0696430206298828,
"learning_rate": 0.003,
"loss": 3.2075405883789063,
"step": 81300
},
{
"epoch": 0.1778209829852189,
"grad_norm": 0.1451922506093979,
"learning_rate": 0.003,
"loss": 3.1747747802734376,
"step": 81400
},
{
"epoch": 0.17803943628126953,
"grad_norm": 0.1978912055492401,
"learning_rate": 0.003,
"loss": 3.223821716308594,
"step": 81500
},
{
"epoch": 0.17825788957732017,
"grad_norm": 0.1773586869239807,
"learning_rate": 0.003,
"loss": 3.2137265014648437,
"step": 81600
},
{
"epoch": 0.17847634287337083,
"grad_norm": 0.15452352166175842,
"learning_rate": 0.003,
"loss": 3.19757568359375,
"step": 81700
},
{
"epoch": 0.17869479616942147,
"grad_norm": 0.16110122203826904,
"learning_rate": 0.003,
"loss": 3.2170181274414062,
"step": 81800
},
{
"epoch": 0.1789132494654721,
"grad_norm": 0.2923918068408966,
"learning_rate": 0.003,
"loss": 3.2086199951171874,
"step": 81900
},
{
"epoch": 0.17913170276152274,
"grad_norm": 0.13742706179618835,
"learning_rate": 0.003,
"loss": 3.203441467285156,
"step": 82000
},
{
"epoch": 0.17935015605757337,
"grad_norm": 0.2099335640668869,
"learning_rate": 0.003,
"loss": 3.18805908203125,
"step": 82100
},
{
"epoch": 0.179568609353624,
"grad_norm": 0.17757396399974823,
"learning_rate": 0.003,
"loss": 3.237929382324219,
"step": 82200
},
{
"epoch": 0.17978706264967464,
"grad_norm": 0.1552405059337616,
"learning_rate": 0.003,
"loss": 3.2098583984375,
"step": 82300
},
{
"epoch": 0.18000551594572528,
"grad_norm": 0.18570825457572937,
"learning_rate": 0.003,
"loss": 3.218511047363281,
"step": 82400
},
{
"epoch": 0.18022396924177592,
"grad_norm": 0.17937444150447845,
"learning_rate": 0.003,
"loss": 3.1134356689453124,
"step": 82500
},
{
"epoch": 0.18044242253782655,
"grad_norm": 0.2126537710428238,
"learning_rate": 0.003,
"loss": 3.181529235839844,
"step": 82600
},
{
"epoch": 0.1806608758338772,
"grad_norm": 0.21367910504341125,
"learning_rate": 0.003,
"loss": 3.201230773925781,
"step": 82700
},
{
"epoch": 0.18087932912992782,
"grad_norm": 0.14153313636779785,
"learning_rate": 0.003,
"loss": 3.1745318603515624,
"step": 82800
},
{
"epoch": 0.18109778242597846,
"grad_norm": 0.19468209147453308,
"learning_rate": 0.003,
"loss": 3.2247454833984377,
"step": 82900
},
{
"epoch": 0.1813162357220291,
"grad_norm": 0.15372858941555023,
"learning_rate": 0.003,
"loss": 3.1692236328125,
"step": 83000
},
{
"epoch": 0.18153468901807973,
"grad_norm": 0.26405900716781616,
"learning_rate": 0.003,
"loss": 3.2481961059570312,
"step": 83100
},
{
"epoch": 0.18175314231413037,
"grad_norm": 0.13953185081481934,
"learning_rate": 0.003,
"loss": 3.185765686035156,
"step": 83200
},
{
"epoch": 0.181971595610181,
"grad_norm": 0.17543235421180725,
"learning_rate": 0.003,
"loss": 3.2108160400390626,
"step": 83300
},
{
"epoch": 0.18219004890623167,
"grad_norm": 0.1657751351594925,
"learning_rate": 0.003,
"loss": 3.248452453613281,
"step": 83400
},
{
"epoch": 0.1824085022022823,
"grad_norm": 0.169894739985466,
"learning_rate": 0.003,
"loss": 3.228109130859375,
"step": 83500
},
{
"epoch": 0.18262695549833294,
"grad_norm": 0.3021637499332428,
"learning_rate": 0.003,
"loss": 3.2393258666992186,
"step": 83600
},
{
"epoch": 0.18284540879438357,
"grad_norm": 0.15600207448005676,
"learning_rate": 0.003,
"loss": 3.2102398681640625,
"step": 83700
},
{
"epoch": 0.1830638620904342,
"grad_norm": 0.140744149684906,
"learning_rate": 0.003,
"loss": 3.1667059326171874,
"step": 83800
},
{
"epoch": 0.18328231538648485,
"grad_norm": 0.15589119493961334,
"learning_rate": 0.003,
"loss": 3.1936209106445315,
"step": 83900
},
{
"epoch": 0.18350076868253548,
"grad_norm": 0.1463201940059662,
"learning_rate": 0.003,
"loss": 3.175903015136719,
"step": 84000
},
{
"epoch": 0.18371922197858612,
"grad_norm": 0.2008126974105835,
"learning_rate": 0.003,
"loss": 3.1910275268554686,
"step": 84100
},
{
"epoch": 0.18393767527463675,
"grad_norm": 0.17553767561912537,
"learning_rate": 0.003,
"loss": 3.187009582519531,
"step": 84200
},
{
"epoch": 0.1841561285706874,
"grad_norm": 0.16969048976898193,
"learning_rate": 0.003,
"loss": 3.21021240234375,
"step": 84300
},
{
"epoch": 0.18437458186673802,
"grad_norm": 0.5377168655395508,
"learning_rate": 0.003,
"loss": 3.1883334350585937,
"step": 84400
},
{
"epoch": 0.18459303516278866,
"grad_norm": 0.1619803011417389,
"learning_rate": 0.003,
"loss": 3.1760284423828127,
"step": 84500
},
{
"epoch": 0.1848114884588393,
"grad_norm": 0.16926124691963196,
"learning_rate": 0.003,
"loss": 3.1831094360351564,
"step": 84600
},
{
"epoch": 0.18502994175488993,
"grad_norm": 0.16700851917266846,
"learning_rate": 0.003,
"loss": 3.1721905517578124,
"step": 84700
},
{
"epoch": 0.18524839505094057,
"grad_norm": 0.14603792130947113,
"learning_rate": 0.003,
"loss": 3.16665283203125,
"step": 84800
},
{
"epoch": 0.1854668483469912,
"grad_norm": 0.24774129688739777,
"learning_rate": 0.003,
"loss": 2.9768441772460936,
"step": 84900
},
{
"epoch": 0.18568530164304184,
"grad_norm": 0.17643874883651733,
"learning_rate": 0.003,
"loss": 2.7105810546875,
"step": 85000
},
{
"epoch": 0.1859037549390925,
"grad_norm": 0.16816070675849915,
"learning_rate": 0.003,
"loss": 3.1750653076171873,
"step": 85100
},
{
"epoch": 0.18612220823514314,
"grad_norm": 0.35066846013069153,
"learning_rate": 0.003,
"loss": 3.204986572265625,
"step": 85200
},
{
"epoch": 0.18634066153119377,
"grad_norm": 0.17999179661273956,
"learning_rate": 0.003,
"loss": 3.170851135253906,
"step": 85300
},
{
"epoch": 0.1865591148272444,
"grad_norm": 0.14632770419120789,
"learning_rate": 0.003,
"loss": 3.1513735961914064,
"step": 85400
},
{
"epoch": 0.18677756812329505,
"grad_norm": 0.14943662285804749,
"learning_rate": 0.003,
"loss": 3.226905517578125,
"step": 85500
},
{
"epoch": 0.18699602141934568,
"grad_norm": 0.1464826613664627,
"learning_rate": 0.003,
"loss": 3.183929138183594,
"step": 85600
},
{
"epoch": 0.18721447471539632,
"grad_norm": 0.1445285826921463,
"learning_rate": 0.003,
"loss": 3.185249938964844,
"step": 85700
},
{
"epoch": 0.18743292801144695,
"grad_norm": 0.15161272883415222,
"learning_rate": 0.003,
"loss": 3.1968368530273437,
"step": 85800
},
{
"epoch": 0.1876513813074976,
"grad_norm": 0.14321133494377136,
"learning_rate": 0.003,
"loss": 3.16796875,
"step": 85900
},
{
"epoch": 0.18786983460354822,
"grad_norm": 0.17690575122833252,
"learning_rate": 0.003,
"loss": 3.079221496582031,
"step": 86000
},
{
"epoch": 0.18808828789959886,
"grad_norm": 0.30789339542388916,
"learning_rate": 0.003,
"loss": 3.1597393798828124,
"step": 86100
},
{
"epoch": 0.1883067411956495,
"grad_norm": 0.19097469747066498,
"learning_rate": 0.003,
"loss": 3.105707092285156,
"step": 86200
},
{
"epoch": 0.18852519449170013,
"grad_norm": 0.1602492481470108,
"learning_rate": 0.003,
"loss": 3.1880517578125,
"step": 86300
},
{
"epoch": 0.18874364778775077,
"grad_norm": 0.1806216835975647,
"learning_rate": 0.003,
"loss": 3.1760809326171877,
"step": 86400
},
{
"epoch": 0.1889621010838014,
"grad_norm": 0.1637730896472931,
"learning_rate": 0.003,
"loss": 3.1409722900390626,
"step": 86500
},
{
"epoch": 0.18918055437985204,
"grad_norm": 0.2023204267024994,
"learning_rate": 0.003,
"loss": 3.164836120605469,
"step": 86600
},
{
"epoch": 0.1893990076759027,
"grad_norm": 0.1658940315246582,
"learning_rate": 0.003,
"loss": 3.201011047363281,
"step": 86700
},
{
"epoch": 0.18961746097195334,
"grad_norm": 0.1435602456331253,
"learning_rate": 0.003,
"loss": 3.1786843872070314,
"step": 86800
},
{
"epoch": 0.18983591426800397,
"grad_norm": 0.18763457238674164,
"learning_rate": 0.003,
"loss": 3.206578674316406,
"step": 86900
},
{
"epoch": 0.1900543675640546,
"grad_norm": 0.25697416067123413,
"learning_rate": 0.003,
"loss": 3.2007901000976564,
"step": 87000
},
{
"epoch": 0.19027282086010525,
"grad_norm": 0.20490820705890656,
"learning_rate": 0.003,
"loss": 3.161679382324219,
"step": 87100
},
{
"epoch": 0.19049127415615588,
"grad_norm": 0.18808764219284058,
"learning_rate": 0.003,
"loss": 3.2807861328125,
"step": 87200
},
{
"epoch": 0.19070972745220652,
"grad_norm": 0.20605118572711945,
"learning_rate": 0.003,
"loss": 3.184476623535156,
"step": 87300
},
{
"epoch": 0.19092818074825715,
"grad_norm": 0.15393364429473877,
"learning_rate": 0.003,
"loss": 3.179691162109375,
"step": 87400
},
{
"epoch": 0.1911466340443078,
"grad_norm": 0.16213102638721466,
"learning_rate": 0.003,
"loss": 3.1681033325195314,
"step": 87500
},
{
"epoch": 0.19136508734035843,
"grad_norm": 0.1807558238506317,
"learning_rate": 0.003,
"loss": 3.190770263671875,
"step": 87600
},
{
"epoch": 0.19158354063640906,
"grad_norm": 0.15362854301929474,
"learning_rate": 0.003,
"loss": 3.1951593017578124,
"step": 87700
},
{
"epoch": 0.1918019939324597,
"grad_norm": 0.3159952163696289,
"learning_rate": 0.003,
"loss": 3.156473388671875,
"step": 87800
},
{
"epoch": 0.19202044722851033,
"grad_norm": 0.2071072906255722,
"learning_rate": 0.003,
"loss": 3.1962716674804685,
"step": 87900
},
{
"epoch": 0.19223890052456097,
"grad_norm": 0.21645040810108185,
"learning_rate": 0.003,
"loss": 2.6888922119140624,
"step": 88000
},
{
"epoch": 0.1924573538206116,
"grad_norm": 0.14421090483665466,
"learning_rate": 0.003,
"loss": 2.664625244140625,
"step": 88100
},
{
"epoch": 0.19267580711666224,
"grad_norm": 0.14521977305412292,
"learning_rate": 0.003,
"loss": 3.1772064208984374,
"step": 88200
},
{
"epoch": 0.19289426041271288,
"grad_norm": 0.18281616270542145,
"learning_rate": 0.003,
"loss": 3.154764404296875,
"step": 88300
},
{
"epoch": 0.19311271370876354,
"grad_norm": 0.16759873926639557,
"learning_rate": 0.003,
"loss": 3.2332708740234377,
"step": 88400
},
{
"epoch": 0.19333116700481418,
"grad_norm": 0.1650831699371338,
"learning_rate": 0.003,
"loss": 3.1961322021484375,
"step": 88500
},
{
"epoch": 0.1935496203008648,
"grad_norm": 0.15335407853126526,
"learning_rate": 0.003,
"loss": 3.219986572265625,
"step": 88600
},
{
"epoch": 0.19376807359691545,
"grad_norm": 0.17449705302715302,
"learning_rate": 0.003,
"loss": 3.1900616455078126,
"step": 88700
},
{
"epoch": 0.19398652689296608,
"grad_norm": 0.17153072357177734,
"learning_rate": 0.003,
"loss": 3.1804873657226564,
"step": 88800
},
{
"epoch": 0.19420498018901672,
"grad_norm": 0.2380538284778595,
"learning_rate": 0.003,
"loss": 3.2428970336914062,
"step": 88900
},
{
"epoch": 0.19442343348506735,
"grad_norm": 0.17524601519107819,
"learning_rate": 0.003,
"loss": 3.185479736328125,
"step": 89000
},
{
"epoch": 0.194641886781118,
"grad_norm": 0.18849943578243256,
"learning_rate": 0.003,
"loss": 3.21683837890625,
"step": 89100
},
{
"epoch": 0.19486034007716863,
"grad_norm": 0.22329972684383392,
"learning_rate": 0.003,
"loss": 3.213042907714844,
"step": 89200
},
{
"epoch": 0.19507879337321926,
"grad_norm": 0.18111512064933777,
"learning_rate": 0.003,
"loss": 3.1977069091796877,
"step": 89300
},
{
"epoch": 0.1952972466692699,
"grad_norm": 0.16808584332466125,
"learning_rate": 0.003,
"loss": 3.2011788940429686,
"step": 89400
},
{
"epoch": 0.19551569996532053,
"grad_norm": 0.16412976384162903,
"learning_rate": 0.003,
"loss": 3.1853192138671873,
"step": 89500
},
{
"epoch": 0.19573415326137117,
"grad_norm": 0.14178995788097382,
"learning_rate": 0.003,
"loss": 3.1789019775390623,
"step": 89600
},
{
"epoch": 0.1959526065574218,
"grad_norm": 0.16296924650669098,
"learning_rate": 0.003,
"loss": 3.190277099609375,
"step": 89700
},
{
"epoch": 0.19617105985347244,
"grad_norm": 0.15710635483264923,
"learning_rate": 0.003,
"loss": 3.206595764160156,
"step": 89800
},
{
"epoch": 0.19638951314952308,
"grad_norm": 0.16058866679668427,
"learning_rate": 0.003,
"loss": 3.1978875732421876,
"step": 89900
},
{
"epoch": 0.19660796644557374,
"grad_norm": 0.19483326375484467,
"learning_rate": 0.003,
"loss": 3.167970886230469,
"step": 90000
},
{
"epoch": 0.19682641974162438,
"grad_norm": 0.24953538179397583,
"learning_rate": 0.003,
"loss": 3.189050598144531,
"step": 90100
},
{
"epoch": 0.197044873037675,
"grad_norm": 0.17547492682933807,
"learning_rate": 0.003,
"loss": 3.175340270996094,
"step": 90200
},
{
"epoch": 0.19726332633372565,
"grad_norm": 0.1600673347711563,
"learning_rate": 0.003,
"loss": 3.142608947753906,
"step": 90300
},
{
"epoch": 0.19748177962977628,
"grad_norm": 0.19891978800296783,
"learning_rate": 0.003,
"loss": 3.2025765991210937,
"step": 90400
},
{
"epoch": 0.19770023292582692,
"grad_norm": 0.21992084383964539,
"learning_rate": 0.003,
"loss": 3.1964657592773436,
"step": 90500
},
{
"epoch": 0.19791868622187755,
"grad_norm": 0.19522127509117126,
"learning_rate": 0.003,
"loss": 3.195756530761719,
"step": 90600
},
{
"epoch": 0.1981371395179282,
"grad_norm": 0.18746817111968994,
"learning_rate": 0.003,
"loss": 3.184117736816406,
"step": 90700
},
{
"epoch": 0.19835559281397883,
"grad_norm": 0.15298475325107574,
"learning_rate": 0.003,
"loss": 3.183582458496094,
"step": 90800
},
{
"epoch": 0.19857404611002946,
"grad_norm": 0.19623512029647827,
"learning_rate": 0.003,
"loss": 3.187120361328125,
"step": 90900
},
{
"epoch": 0.1987924994060801,
"grad_norm": 0.17852617800235748,
"learning_rate": 0.003,
"loss": 3.1580520629882813,
"step": 91000
},
{
"epoch": 0.19901095270213073,
"grad_norm": 0.22779320180416107,
"learning_rate": 0.003,
"loss": 3.1618084716796875,
"step": 91100
},
{
"epoch": 0.19922940599818137,
"grad_norm": 0.24420838057994843,
"learning_rate": 0.003,
"loss": 3.189275817871094,
"step": 91200
},
{
"epoch": 0.199447859294232,
"grad_norm": 0.17383147776126862,
"learning_rate": 0.003,
"loss": 3.1706887817382814,
"step": 91300
},
{
"epoch": 0.19966631259028264,
"grad_norm": 0.21523991227149963,
"learning_rate": 0.003,
"loss": 3.20406005859375,
"step": 91400
},
{
"epoch": 0.19988476588633328,
"grad_norm": 0.19733121991157532,
"learning_rate": 0.003,
"loss": 3.18106689453125,
"step": 91500
},
{
"epoch": 0.2001032191823839,
"grad_norm": 0.22286085784435272,
"learning_rate": 0.003,
"loss": 3.1443585205078124,
"step": 91600
},
{
"epoch": 0.20032167247843458,
"grad_norm": 0.25405412912368774,
"learning_rate": 0.003,
"loss": 3.201405029296875,
"step": 91700
},
{
"epoch": 0.2005401257744852,
"grad_norm": 0.18054093420505524,
"learning_rate": 0.003,
"loss": 3.1207098388671874,
"step": 91800
},
{
"epoch": 0.20075857907053585,
"grad_norm": 0.2531103193759918,
"learning_rate": 0.003,
"loss": 3.172463073730469,
"step": 91900
},
{
"epoch": 0.20097703236658648,
"grad_norm": 0.19328059256076813,
"learning_rate": 0.003,
"loss": 3.20841796875,
"step": 92000
},
{
"epoch": 0.20119548566263712,
"grad_norm": 0.16797882318496704,
"learning_rate": 0.003,
"loss": 3.1828482055664065,
"step": 92100
},
{
"epoch": 0.20141393895868775,
"grad_norm": 0.1891237497329712,
"learning_rate": 0.003,
"loss": 3.187022399902344,
"step": 92200
},
{
"epoch": 0.2016323922547384,
"grad_norm": 0.18582101166248322,
"learning_rate": 0.003,
"loss": 3.1867816162109377,
"step": 92300
},
{
"epoch": 0.20185084555078903,
"grad_norm": 0.2223065346479416,
"learning_rate": 0.003,
"loss": 3.1994952392578124,
"step": 92400
},
{
"epoch": 0.20206929884683966,
"grad_norm": 0.13481435179710388,
"learning_rate": 0.003,
"loss": 3.17259033203125,
"step": 92500
},
{
"epoch": 0.2022877521428903,
"grad_norm": 0.1694314032793045,
"learning_rate": 0.003,
"loss": 3.179879455566406,
"step": 92600
},
{
"epoch": 0.20250620543894093,
"grad_norm": 0.14082738757133484,
"learning_rate": 0.003,
"loss": 3.1622433471679687,
"step": 92700
},
{
"epoch": 0.20272465873499157,
"grad_norm": 0.15445734560489655,
"learning_rate": 0.003,
"loss": 3.146681213378906,
"step": 92800
},
{
"epoch": 0.2029431120310422,
"grad_norm": 0.1701645702123642,
"learning_rate": 0.003,
"loss": 3.2219207763671873,
"step": 92900
},
{
"epoch": 0.20316156532709284,
"grad_norm": 0.16084595024585724,
"learning_rate": 0.003,
"loss": 3.171546630859375,
"step": 93000
},
{
"epoch": 0.20338001862314348,
"grad_norm": 0.1549762487411499,
"learning_rate": 0.003,
"loss": 3.1478482055664063,
"step": 93100
},
{
"epoch": 0.2035984719191941,
"grad_norm": 0.15329527854919434,
"learning_rate": 0.003,
"loss": 3.199483642578125,
"step": 93200
},
{
"epoch": 0.20381692521524478,
"grad_norm": 0.1530144065618515,
"learning_rate": 0.003,
"loss": 3.1848526000976562,
"step": 93300
},
{
"epoch": 0.2040353785112954,
"grad_norm": 0.1716870814561844,
"learning_rate": 0.003,
"loss": 3.154272155761719,
"step": 93400
},
{
"epoch": 0.20425383180734605,
"grad_norm": 0.16520437598228455,
"learning_rate": 0.003,
"loss": 3.232624206542969,
"step": 93500
},
{
"epoch": 0.20447228510339668,
"grad_norm": 0.22018638253211975,
"learning_rate": 0.003,
"loss": 3.177936706542969,
"step": 93600
},
{
"epoch": 0.20469073839944732,
"grad_norm": 0.18284861743450165,
"learning_rate": 0.003,
"loss": 3.1810137939453127,
"step": 93700
},
{
"epoch": 0.20490919169549796,
"grad_norm": 0.22632446885108948,
"learning_rate": 0.003,
"loss": 3.1926318359375,
"step": 93800
},
{
"epoch": 0.2051276449915486,
"grad_norm": 0.16929064691066742,
"learning_rate": 0.003,
"loss": 3.1332843017578127,
"step": 93900
},
{
"epoch": 0.20534609828759923,
"grad_norm": 0.15234090387821198,
"learning_rate": 0.003,
"loss": 3.1795526123046876,
"step": 94000
},
{
"epoch": 0.20556455158364986,
"grad_norm": 1.2888211011886597,
"learning_rate": 0.003,
"loss": 3.1836972045898437,
"step": 94100
},
{
"epoch": 0.2057830048797005,
"grad_norm": 0.15114560723304749,
"learning_rate": 0.003,
"loss": 3.171034240722656,
"step": 94200
},
{
"epoch": 0.20600145817575113,
"grad_norm": 0.22760440409183502,
"learning_rate": 0.003,
"loss": 3.1925299072265627,
"step": 94300
},
{
"epoch": 0.20621991147180177,
"grad_norm": 0.1724519431591034,
"learning_rate": 0.003,
"loss": 3.1574478149414062,
"step": 94400
},
{
"epoch": 0.2064383647678524,
"grad_norm": 0.14399363100528717,
"learning_rate": 0.003,
"loss": 3.160289306640625,
"step": 94500
},
{
"epoch": 0.20665681806390304,
"grad_norm": 0.21455234289169312,
"learning_rate": 0.003,
"loss": 3.2192218017578127,
"step": 94600
},
{
"epoch": 0.20687527135995368,
"grad_norm": 0.143798366189003,
"learning_rate": 0.003,
"loss": 3.2235052490234377,
"step": 94700
},
{
"epoch": 0.2070937246560043,
"grad_norm": 0.13446617126464844,
"learning_rate": 0.003,
"loss": 3.166502685546875,
"step": 94800
},
{
"epoch": 0.20731217795205495,
"grad_norm": 0.1853928416967392,
"learning_rate": 0.003,
"loss": 3.205904846191406,
"step": 94900
},
{
"epoch": 0.2075306312481056,
"grad_norm": 0.1520940512418747,
"learning_rate": 0.003,
"loss": 3.2054425048828126,
"step": 95000
},
{
"epoch": 0.20774908454415625,
"grad_norm": 0.18492989242076874,
"learning_rate": 0.003,
"loss": 3.1887503051757813,
"step": 95100
},
{
"epoch": 0.20796753784020688,
"grad_norm": 0.16937538981437683,
"learning_rate": 0.003,
"loss": 3.1389309692382814,
"step": 95200
},
{
"epoch": 0.20818599113625752,
"grad_norm": 0.1527348905801773,
"learning_rate": 0.003,
"loss": 3.2123422241210937,
"step": 95300
},
{
"epoch": 0.20840444443230816,
"grad_norm": 0.236098974943161,
"learning_rate": 0.003,
"loss": 3.239290466308594,
"step": 95400
},
{
"epoch": 0.2086228977283588,
"grad_norm": 0.14385886490345,
"learning_rate": 0.003,
"loss": 3.211983947753906,
"step": 95500
},
{
"epoch": 0.20884135102440943,
"grad_norm": 0.15831750631332397,
"learning_rate": 0.003,
"loss": 3.222947082519531,
"step": 95600
},
{
"epoch": 0.20905980432046006,
"grad_norm": 0.1465398520231247,
"learning_rate": 0.003,
"loss": 3.2200616455078124,
"step": 95700
},
{
"epoch": 0.2092782576165107,
"grad_norm": 0.2297724485397339,
"learning_rate": 0.003,
"loss": 3.1978054809570313,
"step": 95800
},
{
"epoch": 0.20949671091256133,
"grad_norm": 0.19791044294834137,
"learning_rate": 0.003,
"loss": 3.21572021484375,
"step": 95900
},
{
"epoch": 0.20971516420861197,
"grad_norm": 0.16529631614685059,
"learning_rate": 0.003,
"loss": 3.2116912841796874,
"step": 96000
},
{
"epoch": 0.2099336175046626,
"grad_norm": 0.1494702845811844,
"learning_rate": 0.003,
"loss": 3.1520919799804688,
"step": 96100
},
{
"epoch": 0.21015207080071324,
"grad_norm": 0.18800431489944458,
"learning_rate": 0.003,
"loss": 3.1841400146484373,
"step": 96200
},
{
"epoch": 0.21037052409676388,
"grad_norm": 0.16817855834960938,
"learning_rate": 0.003,
"loss": 3.1894522094726563,
"step": 96300
},
{
"epoch": 0.2105889773928145,
"grad_norm": 0.14505019783973694,
"learning_rate": 0.003,
"loss": 3.1698458862304686,
"step": 96400
},
{
"epoch": 0.21080743068886515,
"grad_norm": 0.3214562237262726,
"learning_rate": 0.003,
"loss": 3.2179776000976563,
"step": 96500
},
{
"epoch": 0.2110258839849158,
"grad_norm": 0.2016151249408722,
"learning_rate": 0.003,
"loss": 3.217587890625,
"step": 96600
},
{
"epoch": 0.21124433728096645,
"grad_norm": 0.1868591159582138,
"learning_rate": 0.003,
"loss": 3.248324279785156,
"step": 96700
},
{
"epoch": 0.21146279057701708,
"grad_norm": 0.14070992171764374,
"learning_rate": 0.003,
"loss": 3.188021545410156,
"step": 96800
},
{
"epoch": 0.21168124387306772,
"grad_norm": 0.20569053292274475,
"learning_rate": 0.003,
"loss": 3.179546203613281,
"step": 96900
},
{
"epoch": 0.21189969716911836,
"grad_norm": 0.15633311867713928,
"learning_rate": 0.003,
"loss": 2.98917724609375,
"step": 97000
},
{
"epoch": 0.212118150465169,
"grad_norm": 0.13893213868141174,
"learning_rate": 0.003,
"loss": 3.2265609741210937,
"step": 97100
},
{
"epoch": 0.21233660376121963,
"grad_norm": 0.13560397922992706,
"learning_rate": 0.003,
"loss": 3.161188049316406,
"step": 97200
},
{
"epoch": 0.21255505705727026,
"grad_norm": 0.22136159241199493,
"learning_rate": 0.003,
"loss": 3.1796783447265624,
"step": 97300
},
{
"epoch": 0.2127735103533209,
"grad_norm": 0.20529834926128387,
"learning_rate": 0.003,
"loss": 3.184031677246094,
"step": 97400
},
{
"epoch": 0.21299196364937154,
"grad_norm": 0.16357794404029846,
"learning_rate": 0.003,
"loss": 3.205530090332031,
"step": 97500
},
{
"epoch": 0.21321041694542217,
"grad_norm": 0.1573294699192047,
"learning_rate": 0.003,
"loss": 3.1988009643554687,
"step": 97600
},
{
"epoch": 0.2134288702414728,
"grad_norm": 0.16560779511928558,
"learning_rate": 0.003,
"loss": 3.24091796875,
"step": 97700
},
{
"epoch": 0.21364732353752344,
"grad_norm": 0.1558590829372406,
"learning_rate": 0.003,
"loss": 3.2470822143554687,
"step": 97800
},
{
"epoch": 0.21386577683357408,
"grad_norm": 0.13487383723258972,
"learning_rate": 0.003,
"loss": 3.21383544921875,
"step": 97900
},
{
"epoch": 0.21408423012962471,
"grad_norm": 0.15421494841575623,
"learning_rate": 0.003,
"loss": 3.231658935546875,
"step": 98000
},
{
"epoch": 0.21430268342567535,
"grad_norm": 0.17839591205120087,
"learning_rate": 0.003,
"loss": 3.2164346313476564,
"step": 98100
},
{
"epoch": 0.21452113672172599,
"grad_norm": 0.1539018750190735,
"learning_rate": 0.003,
"loss": 3.2122412109375,
"step": 98200
},
{
"epoch": 0.21473959001777665,
"grad_norm": 0.16570475697517395,
"learning_rate": 0.003,
"loss": 3.2146701049804687,
"step": 98300
},
{
"epoch": 0.21495804331382729,
"grad_norm": 0.17220532894134521,
"learning_rate": 0.003,
"loss": 3.193248291015625,
"step": 98400
},
{
"epoch": 0.21517649660987792,
"grad_norm": 0.1638244241476059,
"learning_rate": 0.003,
"loss": 3.2288504028320313,
"step": 98500
},
{
"epoch": 0.21539494990592856,
"grad_norm": 0.19310399889945984,
"learning_rate": 0.003,
"loss": 3.20605712890625,
"step": 98600
},
{
"epoch": 0.2156134032019792,
"grad_norm": 0.16728836297988892,
"learning_rate": 0.003,
"loss": 3.182777099609375,
"step": 98700
},
{
"epoch": 0.21583185649802983,
"grad_norm": 0.15934403240680695,
"learning_rate": 0.003,
"loss": 3.194869079589844,
"step": 98800
},
{
"epoch": 0.21605030979408046,
"grad_norm": 0.16646726429462433,
"learning_rate": 0.003,
"loss": 3.207718505859375,
"step": 98900
},
{
"epoch": 0.2162687630901311,
"grad_norm": 0.2790140211582184,
"learning_rate": 0.003,
"loss": 3.1770614624023437,
"step": 99000
},
{
"epoch": 0.21648721638618174,
"grad_norm": 0.1910792887210846,
"learning_rate": 0.003,
"loss": 3.189217529296875,
"step": 99100
},
{
"epoch": 0.21670566968223237,
"grad_norm": 0.19499088823795319,
"learning_rate": 0.003,
"loss": 3.2451211547851564,
"step": 99200
},
{
"epoch": 0.216924122978283,
"grad_norm": 0.13704334199428558,
"learning_rate": 0.003,
"loss": 3.232183532714844,
"step": 99300
},
{
"epoch": 0.21714257627433364,
"grad_norm": 0.1589747965335846,
"learning_rate": 0.003,
"loss": 3.2211962890625,
"step": 99400
},
{
"epoch": 0.21736102957038428,
"grad_norm": 0.1402156502008438,
"learning_rate": 0.003,
"loss": 3.2198226928710936,
"step": 99500
},
{
"epoch": 0.21757948286643491,
"grad_norm": 0.15935233235359192,
"learning_rate": 0.003,
"loss": 3.2299127197265625,
"step": 99600
},
{
"epoch": 0.21779793616248555,
"grad_norm": 0.15546433627605438,
"learning_rate": 0.003,
"loss": 3.202350769042969,
"step": 99700
},
{
"epoch": 0.21801638945853619,
"grad_norm": 0.20298254489898682,
"learning_rate": 0.003,
"loss": 3.221640930175781,
"step": 99800
},
{
"epoch": 0.21823484275458685,
"grad_norm": 0.18505309522151947,
"learning_rate": 0.003,
"loss": 3.217315673828125,
"step": 99900
},
{
"epoch": 0.21845329605063749,
"grad_norm": 0.2007245421409607,
"learning_rate": 0.003,
"loss": 3.23029052734375,
"step": 100000
},
{
"epoch": 0.21867174934668812,
"grad_norm": 0.16363219916820526,
"learning_rate": 0.003,
"loss": 3.2147122192382813,
"step": 100100
},
{
"epoch": 0.21889020264273876,
"grad_norm": 0.19433718919754028,
"learning_rate": 0.003,
"loss": 3.2266812133789062,
"step": 100200
},
{
"epoch": 0.2191086559387894,
"grad_norm": 0.14650124311447144,
"learning_rate": 0.003,
"loss": 3.198697509765625,
"step": 100300
},
{
"epoch": 0.21932710923484003,
"grad_norm": 0.14737825095653534,
"learning_rate": 0.003,
"loss": 3.178011779785156,
"step": 100400
},
{
"epoch": 0.21954556253089066,
"grad_norm": 0.19723723828792572,
"learning_rate": 0.003,
"loss": 3.2108416748046875,
"step": 100500
},
{
"epoch": 0.2197640158269413,
"grad_norm": 0.24265453219413757,
"learning_rate": 0.003,
"loss": 3.247611389160156,
"step": 100600
},
{
"epoch": 0.21998246912299194,
"grad_norm": 0.2030978798866272,
"learning_rate": 0.003,
"loss": 2.7322259521484376,
"step": 100700
},
{
"epoch": 0.22020092241904257,
"grad_norm": 0.1545838713645935,
"learning_rate": 0.003,
"loss": 3.2058575439453123,
"step": 100800
},
{
"epoch": 0.2204193757150932,
"grad_norm": 0.15992237627506256,
"learning_rate": 0.003,
"loss": 3.2379156494140626,
"step": 100900
},
{
"epoch": 0.22063782901114384,
"grad_norm": 0.1919916272163391,
"learning_rate": 0.003,
"loss": 3.2084530639648436,
"step": 101000
},
{
"epoch": 0.22085628230719448,
"grad_norm": 0.16147805750370026,
"learning_rate": 0.003,
"loss": 3.2007907104492186,
"step": 101100
},
{
"epoch": 0.22107473560324511,
"grad_norm": 0.16603071987628937,
"learning_rate": 0.003,
"loss": 3.176233215332031,
"step": 101200
},
{
"epoch": 0.22129318889929575,
"grad_norm": 0.25442710518836975,
"learning_rate": 0.003,
"loss": 3.221510009765625,
"step": 101300
},
{
"epoch": 0.2215116421953464,
"grad_norm": 0.18280309438705444,
"learning_rate": 0.003,
"loss": 3.2156817626953127,
"step": 101400
},
{
"epoch": 0.22173009549139702,
"grad_norm": 0.20255111157894135,
"learning_rate": 0.003,
"loss": 3.22599609375,
"step": 101500
},
{
"epoch": 0.22194854878744769,
"grad_norm": 0.19555102288722992,
"learning_rate": 0.003,
"loss": 3.234263610839844,
"step": 101600
},
{
"epoch": 0.22216700208349832,
"grad_norm": 0.16481703519821167,
"learning_rate": 0.003,
"loss": 3.2006109619140624,
"step": 101700
},
{
"epoch": 0.22238545537954896,
"grad_norm": 0.1800691783428192,
"learning_rate": 0.003,
"loss": 3.2422198486328124,
"step": 101800
},
{
"epoch": 0.2226039086755996,
"grad_norm": 0.19977200031280518,
"learning_rate": 0.003,
"loss": 3.214632568359375,
"step": 101900
},
{
"epoch": 0.22282236197165023,
"grad_norm": 0.18122437596321106,
"learning_rate": 0.003,
"loss": 3.1705487060546873,
"step": 102000
},
{
"epoch": 0.22304081526770086,
"grad_norm": 0.17697519063949585,
"learning_rate": 0.003,
"loss": 3.2142709350585936,
"step": 102100
},
{
"epoch": 0.2232592685637515,
"grad_norm": 0.1942490041255951,
"learning_rate": 0.003,
"loss": 3.192371520996094,
"step": 102200
},
{
"epoch": 0.22347772185980214,
"grad_norm": 0.14101845026016235,
"learning_rate": 0.003,
"loss": 3.1763015747070313,
"step": 102300
},
{
"epoch": 0.22369617515585277,
"grad_norm": 0.15464375913143158,
"learning_rate": 0.003,
"loss": 3.2267938232421876,
"step": 102400
},
{
"epoch": 0.2239146284519034,
"grad_norm": 0.15713827311992645,
"learning_rate": 0.003,
"loss": 3.1892242431640625,
"step": 102500
},
{
"epoch": 0.22413308174795404,
"grad_norm": 0.19050081074237823,
"learning_rate": 0.003,
"loss": 3.203402099609375,
"step": 102600
},
{
"epoch": 0.22435153504400468,
"grad_norm": 0.14306189119815826,
"learning_rate": 0.003,
"loss": 3.2019168090820314,
"step": 102700
},
{
"epoch": 0.22456998834005532,
"grad_norm": 0.17052483558654785,
"learning_rate": 0.003,
"loss": 3.191214294433594,
"step": 102800
},
{
"epoch": 0.22478844163610595,
"grad_norm": 0.15815363824367523,
"learning_rate": 0.003,
"loss": 3.1885943603515625,
"step": 102900
},
{
"epoch": 0.2250068949321566,
"grad_norm": 0.19354431331157684,
"learning_rate": 0.003,
"loss": 3.1964739990234374,
"step": 103000
},
{
"epoch": 0.22522534822820722,
"grad_norm": 0.16281826794147491,
"learning_rate": 0.003,
"loss": 3.2205325317382814,
"step": 103100
},
{
"epoch": 0.2254438015242579,
"grad_norm": 0.173141747713089,
"learning_rate": 0.003,
"loss": 3.1859710693359373,
"step": 103200
},
{
"epoch": 0.22566225482030852,
"grad_norm": 0.1682673543691635,
"learning_rate": 0.003,
"loss": 3.196368408203125,
"step": 103300
},
{
"epoch": 0.22588070811635916,
"grad_norm": 0.20099574327468872,
"learning_rate": 0.003,
"loss": 3.1951303100585937,
"step": 103400
},
{
"epoch": 0.2260991614124098,
"grad_norm": 0.18030375242233276,
"learning_rate": 0.003,
"loss": 3.2207177734375,
"step": 103500
},
{
"epoch": 0.22631761470846043,
"grad_norm": 0.21371352672576904,
"learning_rate": 0.003,
"loss": 3.184823303222656,
"step": 103600
},
{
"epoch": 0.22653606800451107,
"grad_norm": 0.17467491328716278,
"learning_rate": 0.003,
"loss": 3.2101016235351563,
"step": 103700
},
{
"epoch": 0.2267545213005617,
"grad_norm": 0.18792425096035004,
"learning_rate": 0.003,
"loss": 3.1528106689453126,
"step": 103800
},
{
"epoch": 0.22697297459661234,
"grad_norm": 0.238530233502388,
"learning_rate": 0.003,
"loss": 3.1868878173828126,
"step": 103900
},
{
"epoch": 0.22719142789266297,
"grad_norm": 0.155843123793602,
"learning_rate": 0.003,
"loss": 3.170243225097656,
"step": 104000
},
{
"epoch": 0.2274098811887136,
"grad_norm": 0.18454401195049286,
"learning_rate": 0.003,
"loss": 3.1931134033203126,
"step": 104100
},
{
"epoch": 0.22762833448476424,
"grad_norm": 0.1799052506685257,
"learning_rate": 0.003,
"loss": 3.1812493896484373,
"step": 104200
},
{
"epoch": 0.22784678778081488,
"grad_norm": 0.21381621062755585,
"learning_rate": 0.003,
"loss": 3.172951965332031,
"step": 104300
},
{
"epoch": 0.22806524107686552,
"grad_norm": 0.13922972977161407,
"learning_rate": 0.003,
"loss": 3.1882305908203126,
"step": 104400
},
{
"epoch": 0.22828369437291615,
"grad_norm": 0.15168224275112152,
"learning_rate": 0.003,
"loss": 3.2093801879882813,
"step": 104500
},
{
"epoch": 0.2285021476689668,
"grad_norm": 0.23901334404945374,
"learning_rate": 0.003,
"loss": 3.165443115234375,
"step": 104600
},
{
"epoch": 0.22872060096501742,
"grad_norm": 0.16107922792434692,
"learning_rate": 0.003,
"loss": 3.2130291748046873,
"step": 104700
},
{
"epoch": 0.22893905426106806,
"grad_norm": 0.8439098596572876,
"learning_rate": 0.003,
"loss": 3.19970947265625,
"step": 104800
},
{
"epoch": 0.22915750755711872,
"grad_norm": 0.15786705911159515,
"learning_rate": 0.003,
"loss": 3.1903897094726563,
"step": 104900
},
{
"epoch": 0.22937596085316936,
"grad_norm": 0.15847380459308624,
"learning_rate": 0.003,
"loss": 3.189295349121094,
"step": 105000
},
{
"epoch": 0.22959441414922,
"grad_norm": 0.32766076922416687,
"learning_rate": 0.003,
"loss": 3.1828533935546877,
"step": 105100
},
{
"epoch": 0.22981286744527063,
"grad_norm": 0.18468038737773895,
"learning_rate": 0.003,
"loss": 3.196568908691406,
"step": 105200
},
{
"epoch": 0.23003132074132127,
"grad_norm": 0.1972939521074295,
"learning_rate": 0.003,
"loss": 3.159031982421875,
"step": 105300
},
{
"epoch": 0.2302497740373719,
"grad_norm": 0.18358947336673737,
"learning_rate": 0.003,
"loss": 3.173365478515625,
"step": 105400
},
{
"epoch": 0.23046822733342254,
"grad_norm": 0.19217951595783234,
"learning_rate": 0.003,
"loss": 3.2317135620117186,
"step": 105500
},
{
"epoch": 0.23068668062947317,
"grad_norm": 0.17112338542938232,
"learning_rate": 0.003,
"loss": 3.205096435546875,
"step": 105600
},
{
"epoch": 0.2309051339255238,
"grad_norm": 0.14539970457553864,
"learning_rate": 0.003,
"loss": 3.1683807373046875,
"step": 105700
},
{
"epoch": 0.23112358722157444,
"grad_norm": 0.15848980844020844,
"learning_rate": 0.003,
"loss": 3.219859313964844,
"step": 105800
},
{
"epoch": 0.23134204051762508,
"grad_norm": 0.20147928595542908,
"learning_rate": 0.003,
"loss": 3.177248840332031,
"step": 105900
},
{
"epoch": 0.23156049381367572,
"grad_norm": 0.1637161672115326,
"learning_rate": 0.003,
"loss": 3.2011764526367186,
"step": 106000
},
{
"epoch": 0.23177894710972635,
"grad_norm": 0.20364221930503845,
"learning_rate": 0.003,
"loss": 3.1972198486328125,
"step": 106100
},
{
"epoch": 0.231997400405777,
"grad_norm": 0.15852341055870056,
"learning_rate": 0.003,
"loss": 3.1781005859375,
"step": 106200
},
{
"epoch": 0.23221585370182762,
"grad_norm": 0.2659218907356262,
"learning_rate": 0.003,
"loss": 3.2024310302734373,
"step": 106300
},
{
"epoch": 0.23243430699787826,
"grad_norm": 0.15729771554470062,
"learning_rate": 0.003,
"loss": 3.1765545654296874,
"step": 106400
},
{
"epoch": 0.23265276029392892,
"grad_norm": 0.16436509788036346,
"learning_rate": 0.003,
"loss": 3.2103701782226564,
"step": 106500
},
{
"epoch": 0.23287121358997956,
"grad_norm": 0.22796779870986938,
"learning_rate": 0.003,
"loss": 3.211387939453125,
"step": 106600
},
{
"epoch": 0.2330896668860302,
"grad_norm": 0.18208400905132294,
"learning_rate": 0.003,
"loss": 3.1726083374023437,
"step": 106700
},
{
"epoch": 0.23330812018208083,
"grad_norm": 0.18015295267105103,
"learning_rate": 0.003,
"loss": 3.1254708862304685,
"step": 106800
},
{
"epoch": 0.23352657347813147,
"grad_norm": 0.1717136651277542,
"learning_rate": 0.003,
"loss": 3.19731201171875,
"step": 106900
},
{
"epoch": 0.2337450267741821,
"grad_norm": 0.23622432351112366,
"learning_rate": 0.003,
"loss": 3.1919650268554687,
"step": 107000
},
{
"epoch": 0.23396348007023274,
"grad_norm": 0.148057758808136,
"learning_rate": 0.003,
"loss": 3.17125244140625,
"step": 107100
},
{
"epoch": 0.23418193336628337,
"grad_norm": 0.14156973361968994,
"learning_rate": 0.003,
"loss": 3.201607666015625,
"step": 107200
},
{
"epoch": 0.234400386662334,
"grad_norm": 0.3278632164001465,
"learning_rate": 0.003,
"loss": 3.1915264892578126,
"step": 107300
},
{
"epoch": 0.23461883995838465,
"grad_norm": 0.18710091710090637,
"learning_rate": 0.003,
"loss": 3.249506530761719,
"step": 107400
},
{
"epoch": 0.23483729325443528,
"grad_norm": 0.1827370971441269,
"learning_rate": 0.003,
"loss": 3.201341247558594,
"step": 107500
},
{
"epoch": 0.23505574655048592,
"grad_norm": 0.24463097751140594,
"learning_rate": 0.003,
"loss": 3.2041506958007813,
"step": 107600
},
{
"epoch": 0.23527419984653655,
"grad_norm": 0.1820523589849472,
"learning_rate": 0.003,
"loss": 3.2394415283203126,
"step": 107700
},
{
"epoch": 0.2354926531425872,
"grad_norm": 0.17374499142169952,
"learning_rate": 0.003,
"loss": 3.229110107421875,
"step": 107800
},
{
"epoch": 0.23571110643863782,
"grad_norm": 0.1483023315668106,
"learning_rate": 0.003,
"loss": 3.18363037109375,
"step": 107900
},
{
"epoch": 0.23592955973468846,
"grad_norm": 0.17039185762405396,
"learning_rate": 0.003,
"loss": 3.2658441162109373,
"step": 108000
},
{
"epoch": 0.2361480130307391,
"grad_norm": 0.1909257471561432,
"learning_rate": 0.003,
"loss": 3.2034756469726564,
"step": 108100
},
{
"epoch": 0.23636646632678976,
"grad_norm": 0.1725386679172516,
"learning_rate": 0.003,
"loss": 3.2072048950195313,
"step": 108200
},
{
"epoch": 0.2365849196228404,
"grad_norm": 0.14467132091522217,
"learning_rate": 0.003,
"loss": 3.1985675048828126,
"step": 108300
},
{
"epoch": 0.23680337291889103,
"grad_norm": 0.1315174400806427,
"learning_rate": 0.003,
"loss": 3.2306289672851562,
"step": 108400
},
{
"epoch": 0.23702182621494167,
"grad_norm": 0.14547792077064514,
"learning_rate": 0.003,
"loss": 3.2040252685546875,
"step": 108500
},
{
"epoch": 0.2372402795109923,
"grad_norm": 0.1825241893529892,
"learning_rate": 0.003,
"loss": 3.1949697875976564,
"step": 108600
},
{
"epoch": 0.23745873280704294,
"grad_norm": 0.1727125644683838,
"learning_rate": 0.003,
"loss": 3.2289556884765624,
"step": 108700
},
{
"epoch": 0.23767718610309357,
"grad_norm": 0.24597737193107605,
"learning_rate": 0.003,
"loss": 3.223450012207031,
"step": 108800
},
{
"epoch": 0.2378956393991442,
"grad_norm": 0.1414637416601181,
"learning_rate": 0.003,
"loss": 3.232257080078125,
"step": 108900
},
{
"epoch": 0.23811409269519485,
"grad_norm": 0.15455834567546844,
"learning_rate": 0.003,
"loss": 3.2108209228515623,
"step": 109000
},
{
"epoch": 0.23833254599124548,
"grad_norm": 0.2669357359409332,
"learning_rate": 0.003,
"loss": 3.1880422973632814,
"step": 109100
},
{
"epoch": 0.23855099928729612,
"grad_norm": 0.19473835825920105,
"learning_rate": 0.003,
"loss": 3.1808990478515624,
"step": 109200
},
{
"epoch": 0.23876945258334675,
"grad_norm": 0.17744936048984528,
"learning_rate": 0.003,
"loss": 3.1810150146484375,
"step": 109300
},
{
"epoch": 0.2389879058793974,
"grad_norm": 0.1516321897506714,
"learning_rate": 0.003,
"loss": 3.1612823486328123,
"step": 109400
},
{
"epoch": 0.23920635917544802,
"grad_norm": 0.2026938945055008,
"learning_rate": 0.003,
"loss": 3.2155911254882814,
"step": 109500
},
{
"epoch": 0.23942481247149866,
"grad_norm": 0.21379505097866058,
"learning_rate": 0.003,
"loss": 3.2127227783203125,
"step": 109600
},
{
"epoch": 0.2396432657675493,
"grad_norm": 0.13985757529735565,
"learning_rate": 0.003,
"loss": 3.2153839111328124,
"step": 109700
},
{
"epoch": 0.23986171906359996,
"grad_norm": 0.2607848346233368,
"learning_rate": 0.003,
"loss": 3.1715011596679688,
"step": 109800
},
{
"epoch": 0.2400801723596506,
"grad_norm": 0.2337099313735962,
"learning_rate": 0.003,
"loss": 3.1352163696289064,
"step": 109900
},
{
"epoch": 0.24029862565570123,
"grad_norm": 0.1557885706424713,
"learning_rate": 0.003,
"loss": 3.203826904296875,
"step": 110000
},
{
"epoch": 0.24051707895175187,
"grad_norm": 0.15149317681789398,
"learning_rate": 0.003,
"loss": 3.1457794189453123,
"step": 110100
},
{
"epoch": 0.2407355322478025,
"grad_norm": 0.2450362741947174,
"learning_rate": 0.003,
"loss": 3.1606707763671875,
"step": 110200
},
{
"epoch": 0.24095398554385314,
"grad_norm": 0.19675183296203613,
"learning_rate": 0.003,
"loss": 3.136336669921875,
"step": 110300
},
{
"epoch": 0.24117243883990377,
"grad_norm": 0.17948362231254578,
"learning_rate": 0.003,
"loss": 3.158858642578125,
"step": 110400
},
{
"epoch": 0.2413908921359544,
"grad_norm": 0.16643723845481873,
"learning_rate": 0.003,
"loss": 3.157088623046875,
"step": 110500
},
{
"epoch": 0.24160934543200505,
"grad_norm": 0.14619401097297668,
"learning_rate": 0.003,
"loss": 3.132615051269531,
"step": 110600
},
{
"epoch": 0.24182779872805568,
"grad_norm": 0.17994770407676697,
"learning_rate": 0.003,
"loss": 3.145055236816406,
"step": 110700
},
{
"epoch": 0.24204625202410632,
"grad_norm": 0.19818727672100067,
"learning_rate": 0.003,
"loss": 3.173731689453125,
"step": 110800
},
{
"epoch": 0.24226470532015695,
"grad_norm": 0.14523686468601227,
"learning_rate": 0.003,
"loss": 3.1580401611328126,
"step": 110900
},
{
"epoch": 0.2424831586162076,
"grad_norm": 0.16975483298301697,
"learning_rate": 0.003,
"loss": 3.1509103393554687,
"step": 111000
},
{
"epoch": 0.24270161191225822,
"grad_norm": 0.1531265527009964,
"learning_rate": 0.003,
"loss": 3.1599996948242186,
"step": 111100
},
{
"epoch": 0.24292006520830886,
"grad_norm": 0.15926402807235718,
"learning_rate": 0.003,
"loss": 3.142339172363281,
"step": 111200
},
{
"epoch": 0.2431385185043595,
"grad_norm": 0.20109127461910248,
"learning_rate": 0.003,
"loss": 3.1263226318359374,
"step": 111300
},
{
"epoch": 0.24335697180041013,
"grad_norm": 0.19264379143714905,
"learning_rate": 0.003,
"loss": 3.138429870605469,
"step": 111400
},
{
"epoch": 0.2435754250964608,
"grad_norm": 0.2017202228307724,
"learning_rate": 0.003,
"loss": 3.1567633056640627,
"step": 111500
},
{
"epoch": 0.24379387839251143,
"grad_norm": 0.17149969935417175,
"learning_rate": 0.003,
"loss": 3.154873046875,
"step": 111600
},
{
"epoch": 0.24401233168856207,
"grad_norm": 0.1989082247018814,
"learning_rate": 0.003,
"loss": 3.159642333984375,
"step": 111700
},
{
"epoch": 0.2442307849846127,
"grad_norm": 0.20310066640377045,
"learning_rate": 0.003,
"loss": 3.1391470336914065,
"step": 111800
},
{
"epoch": 0.24444923828066334,
"grad_norm": 0.1760169416666031,
"learning_rate": 0.003,
"loss": 3.1731060791015624,
"step": 111900
},
{
"epoch": 0.24466769157671397,
"grad_norm": 0.16488592326641083,
"learning_rate": 0.003,
"loss": 3.17441650390625,
"step": 112000
},
{
"epoch": 0.2448861448727646,
"grad_norm": 0.1397256851196289,
"learning_rate": 0.003,
"loss": 3.062320556640625,
"step": 112100
},
{
"epoch": 0.24510459816881525,
"grad_norm": 0.1884027123451233,
"learning_rate": 0.003,
"loss": 3.170846252441406,
"step": 112200
},
{
"epoch": 0.24532305146486588,
"grad_norm": 0.19250339269638062,
"learning_rate": 0.003,
"loss": 3.136485290527344,
"step": 112300
},
{
"epoch": 0.24554150476091652,
"grad_norm": 0.155894473195076,
"learning_rate": 0.003,
"loss": 3.153933410644531,
"step": 112400
},
{
"epoch": 0.24575995805696715,
"grad_norm": 0.15026205778121948,
"learning_rate": 0.003,
"loss": 3.1478778076171876,
"step": 112500
},
{
"epoch": 0.2459784113530178,
"grad_norm": 0.1964258998632431,
"learning_rate": 0.003,
"loss": 3.173653259277344,
"step": 112600
},
{
"epoch": 0.24619686464906843,
"grad_norm": 0.16881966590881348,
"learning_rate": 0.003,
"loss": 3.2006106567382813,
"step": 112700
},
{
"epoch": 0.24641531794511906,
"grad_norm": 0.189187154173851,
"learning_rate": 0.003,
"loss": 3.164288330078125,
"step": 112800
},
{
"epoch": 0.2466337712411697,
"grad_norm": 0.16848815977573395,
"learning_rate": 0.003,
"loss": 3.1316644287109376,
"step": 112900
},
{
"epoch": 0.24685222453722033,
"grad_norm": 0.15248990058898926,
"learning_rate": 0.003,
"loss": 3.1296035766601564,
"step": 113000
},
{
"epoch": 0.247070677833271,
"grad_norm": 0.21292132139205933,
"learning_rate": 0.003,
"loss": 3.1329031372070313,
"step": 113100
},
{
"epoch": 0.24728913112932163,
"grad_norm": 0.22059094905853271,
"learning_rate": 0.003,
"loss": 3.133268127441406,
"step": 113200
},
{
"epoch": 0.24750758442537227,
"grad_norm": 0.16975048184394836,
"learning_rate": 0.003,
"loss": 3.1503656005859373,
"step": 113300
},
{
"epoch": 0.2477260377214229,
"grad_norm": 0.18195408582687378,
"learning_rate": 0.003,
"loss": 3.1662640380859375,
"step": 113400
},
{
"epoch": 0.24794449101747354,
"grad_norm": 0.16216278076171875,
"learning_rate": 0.003,
"loss": 3.1075955200195313,
"step": 113500
},
{
"epoch": 0.24816294431352418,
"grad_norm": 0.2042403370141983,
"learning_rate": 0.003,
"loss": 3.1465740966796876,
"step": 113600
},
{
"epoch": 0.2483813976095748,
"grad_norm": 0.17982223629951477,
"learning_rate": 0.003,
"loss": 2.918134765625,
"step": 113700
},
{
"epoch": 0.24859985090562545,
"grad_norm": 0.16524387896060944,
"learning_rate": 0.003,
"loss": 3.1114352416992186,
"step": 113800
},
{
"epoch": 0.24881830420167608,
"grad_norm": 0.1808774620294571,
"learning_rate": 0.003,
"loss": 3.1193252563476563,
"step": 113900
},
{
"epoch": 0.24903675749772672,
"grad_norm": 0.18143890798091888,
"learning_rate": 0.003,
"loss": 3.146431579589844,
"step": 114000
},
{
"epoch": 0.24925521079377735,
"grad_norm": 0.1492440104484558,
"learning_rate": 0.003,
"loss": 3.187480773925781,
"step": 114100
},
{
"epoch": 0.249473664089828,
"grad_norm": 0.23453925549983978,
"learning_rate": 0.003,
"loss": 3.08376220703125,
"step": 114200
},
{
"epoch": 0.24969211738587863,
"grad_norm": 0.18290016055107117,
"learning_rate": 0.003,
"loss": 3.118404541015625,
"step": 114300
},
{
"epoch": 0.24991057068192926,
"grad_norm": 0.16905197501182556,
"learning_rate": 0.003,
"loss": 3.1375994873046875,
"step": 114400
},
{
"epoch": 0.2501290239779799,
"grad_norm": 0.17251262068748474,
"learning_rate": 0.003,
"loss": 3.1640499877929686,
"step": 114500
},
{
"epoch": 0.25034747727403056,
"grad_norm": 0.16715526580810547,
"learning_rate": 0.003,
"loss": 3.1812918090820315,
"step": 114600
},
{
"epoch": 0.2505659305700812,
"grad_norm": 0.18836091458797455,
"learning_rate": 0.003,
"loss": 3.099549560546875,
"step": 114700
},
{
"epoch": 0.25078438386613183,
"grad_norm": 0.24706906080245972,
"learning_rate": 0.003,
"loss": 3.1219192504882813,
"step": 114800
},
{
"epoch": 0.25100283716218247,
"grad_norm": 0.16439929604530334,
"learning_rate": 0.003,
"loss": 3.087160339355469,
"step": 114900
},
{
"epoch": 0.2512212904582331,
"grad_norm": 0.21023960411548615,
"learning_rate": 0.003,
"loss": 3.158519287109375,
"step": 115000
},
{
"epoch": 0.25143974375428374,
"grad_norm": 0.15426918864250183,
"learning_rate": 0.003,
"loss": 3.1479229736328125,
"step": 115100
},
{
"epoch": 0.2516581970503344,
"grad_norm": 0.16378799080848694,
"learning_rate": 0.003,
"loss": 3.16879638671875,
"step": 115200
},
{
"epoch": 0.251876650346385,
"grad_norm": 0.15669791400432587,
"learning_rate": 0.003,
"loss": 3.1172305297851564,
"step": 115300
},
{
"epoch": 0.25209510364243565,
"grad_norm": 0.24460077285766602,
"learning_rate": 0.003,
"loss": 3.130740966796875,
"step": 115400
},
{
"epoch": 0.2523135569384863,
"grad_norm": 0.22701039910316467,
"learning_rate": 0.003,
"loss": 3.143014221191406,
"step": 115500
},
{
"epoch": 0.2525320102345369,
"grad_norm": 0.18664608895778656,
"learning_rate": 0.003,
"loss": 3.1169943237304687,
"step": 115600
},
{
"epoch": 0.25275046353058755,
"grad_norm": 0.1917290985584259,
"learning_rate": 0.003,
"loss": 3.132640075683594,
"step": 115700
},
{
"epoch": 0.2529689168266382,
"grad_norm": 0.14543747901916504,
"learning_rate": 0.003,
"loss": 3.174365234375,
"step": 115800
},
{
"epoch": 0.2531873701226888,
"grad_norm": 0.19407229125499725,
"learning_rate": 0.003,
"loss": 3.1486831665039063,
"step": 115900
},
{
"epoch": 0.25340582341873946,
"grad_norm": 0.16081637144088745,
"learning_rate": 0.003,
"loss": 2.9254141235351563,
"step": 116000
},
{
"epoch": 0.2536242767147901,
"grad_norm": 0.20264428853988647,
"learning_rate": 0.003,
"loss": 3.1475180053710936,
"step": 116100
},
{
"epoch": 0.25384273001084073,
"grad_norm": 0.13481463491916656,
"learning_rate": 0.003,
"loss": 3.129048156738281,
"step": 116200
},
{
"epoch": 0.25406118330689137,
"grad_norm": 0.16611292958259583,
"learning_rate": 0.003,
"loss": 3.1350830078125,
"step": 116300
},
{
"epoch": 0.254279636602942,
"grad_norm": 0.18687981367111206,
"learning_rate": 0.003,
"loss": 3.135335693359375,
"step": 116400
},
{
"epoch": 0.25449808989899264,
"grad_norm": 0.2523234188556671,
"learning_rate": 0.003,
"loss": 3.1152474975585935,
"step": 116500
},
{
"epoch": 0.2547165431950433,
"grad_norm": 0.18094249069690704,
"learning_rate": 0.003,
"loss": 3.1234640502929687,
"step": 116600
},
{
"epoch": 0.2549349964910939,
"grad_norm": 0.20491555333137512,
"learning_rate": 0.003,
"loss": 3.1479095458984374,
"step": 116700
},
{
"epoch": 0.25515344978714455,
"grad_norm": 0.20158524811267853,
"learning_rate": 0.003,
"loss": 3.1634478759765625,
"step": 116800
},
{
"epoch": 0.2553719030831952,
"grad_norm": 0.20101554691791534,
"learning_rate": 0.003,
"loss": 3.1396548461914064,
"step": 116900
},
{
"epoch": 0.2555903563792458,
"grad_norm": 0.1438259780406952,
"learning_rate": 0.003,
"loss": 3.1369644165039063,
"step": 117000
},
{
"epoch": 0.25580880967529646,
"grad_norm": 0.1390850841999054,
"learning_rate": 0.003,
"loss": 3.139820251464844,
"step": 117100
},
{
"epoch": 0.25602726297134715,
"grad_norm": 0.1506982147693634,
"learning_rate": 0.003,
"loss": 3.1354669189453124,
"step": 117200
},
{
"epoch": 0.2562457162673978,
"grad_norm": 0.16565769910812378,
"learning_rate": 0.003,
"loss": 3.146010437011719,
"step": 117300
},
{
"epoch": 0.2564641695634484,
"grad_norm": 0.13977010548114777,
"learning_rate": 0.003,
"loss": 3.1299252319335937,
"step": 117400
},
{
"epoch": 0.25668262285949905,
"grad_norm": 0.19699348509311676,
"learning_rate": 0.003,
"loss": 3.121324462890625,
"step": 117500
},
{
"epoch": 0.2569010761555497,
"grad_norm": 0.15574857592582703,
"learning_rate": 0.003,
"loss": 3.154246826171875,
"step": 117600
},
{
"epoch": 0.2571195294516003,
"grad_norm": 0.16224879026412964,
"learning_rate": 0.003,
"loss": 3.1255767822265623,
"step": 117700
},
{
"epoch": 0.25733798274765096,
"grad_norm": 0.19783753156661987,
"learning_rate": 0.003,
"loss": 3.1510748291015624,
"step": 117800
},
{
"epoch": 0.2575564360437016,
"grad_norm": 0.1567201316356659,
"learning_rate": 0.003,
"loss": 3.068137512207031,
"step": 117900
},
{
"epoch": 0.25777488933975223,
"grad_norm": 0.23760242760181427,
"learning_rate": 0.003,
"loss": 3.13155517578125,
"step": 118000
},
{
"epoch": 0.25799334263580287,
"grad_norm": 0.19157196581363678,
"learning_rate": 0.003,
"loss": 3.1604299926757813,
"step": 118100
},
{
"epoch": 0.2582117959318535,
"grad_norm": 0.23139914870262146,
"learning_rate": 0.003,
"loss": 3.1489080810546874,
"step": 118200
},
{
"epoch": 0.25843024922790414,
"grad_norm": 0.1830720454454422,
"learning_rate": 0.003,
"loss": 3.091242370605469,
"step": 118300
},
{
"epoch": 0.2586487025239548,
"grad_norm": 0.20480267703533173,
"learning_rate": 0.003,
"loss": 3.141184997558594,
"step": 118400
},
{
"epoch": 0.2588671558200054,
"grad_norm": 0.16266120970249176,
"learning_rate": 0.003,
"loss": 3.160396728515625,
"step": 118500
},
{
"epoch": 0.25908560911605605,
"grad_norm": 0.18189287185668945,
"learning_rate": 0.003,
"loss": 3.1183380126953124,
"step": 118600
},
{
"epoch": 0.2593040624121067,
"grad_norm": 0.18481861054897308,
"learning_rate": 0.003,
"loss": 3.142648010253906,
"step": 118700
},
{
"epoch": 0.2595225157081573,
"grad_norm": 0.20861801505088806,
"learning_rate": 0.003,
"loss": 3.122877197265625,
"step": 118800
},
{
"epoch": 0.25974096900420796,
"grad_norm": 0.16465097665786743,
"learning_rate": 0.003,
"loss": 3.157867126464844,
"step": 118900
},
{
"epoch": 0.2599594223002586,
"grad_norm": 0.1347455084323883,
"learning_rate": 0.003,
"loss": 3.126204833984375,
"step": 119000
},
{
"epoch": 0.2601778755963092,
"grad_norm": 0.22783678770065308,
"learning_rate": 0.003,
"loss": 3.1145480346679686,
"step": 119100
},
{
"epoch": 0.26039632889235986,
"grad_norm": 0.18908102810382843,
"learning_rate": 0.003,
"loss": 3.1435897827148436,
"step": 119200
},
{
"epoch": 0.2606147821884105,
"grad_norm": 0.2350047528743744,
"learning_rate": 0.003,
"loss": 3.1211563110351563,
"step": 119300
},
{
"epoch": 0.26083323548446113,
"grad_norm": 0.16897456347942352,
"learning_rate": 0.003,
"loss": 3.1367947387695314,
"step": 119400
},
{
"epoch": 0.26105168878051177,
"grad_norm": 0.16670124232769012,
"learning_rate": 0.003,
"loss": 3.121427307128906,
"step": 119500
},
{
"epoch": 0.2612701420765624,
"grad_norm": 0.1661265641450882,
"learning_rate": 0.003,
"loss": 3.096488037109375,
"step": 119600
},
{
"epoch": 0.26148859537261304,
"grad_norm": 0.16746363043785095,
"learning_rate": 0.003,
"loss": 3.1369305419921876,
"step": 119700
},
{
"epoch": 0.2617070486686637,
"grad_norm": 0.16642019152641296,
"learning_rate": 0.003,
"loss": 3.123088073730469,
"step": 119800
},
{
"epoch": 0.2619255019647143,
"grad_norm": 0.352807879447937,
"learning_rate": 0.003,
"loss": 3.11847900390625,
"step": 119900
},
{
"epoch": 0.26214395526076495,
"grad_norm": 0.282635360956192,
"learning_rate": 0.003,
"loss": 3.1500448608398437,
"step": 120000
},
{
"epoch": 0.2623624085568156,
"grad_norm": 0.17607861757278442,
"learning_rate": 0.003,
"loss": 3.1523980712890625,
"step": 120100
},
{
"epoch": 0.2625808618528662,
"grad_norm": 0.20902930200099945,
"learning_rate": 0.003,
"loss": 3.12585205078125,
"step": 120200
},
{
"epoch": 0.26279931514891686,
"grad_norm": 0.16336879134178162,
"learning_rate": 0.003,
"loss": 3.1010125732421874,
"step": 120300
},
{
"epoch": 0.2630177684449675,
"grad_norm": 0.17145244777202606,
"learning_rate": 0.003,
"loss": 3.1193792724609377,
"step": 120400
},
{
"epoch": 0.2632362217410182,
"grad_norm": 0.1874585896730423,
"learning_rate": 0.003,
"loss": 3.112429504394531,
"step": 120500
},
{
"epoch": 0.2634546750370688,
"grad_norm": 0.16599197685718536,
"learning_rate": 0.003,
"loss": 3.1309783935546873,
"step": 120600
},
{
"epoch": 0.26367312833311946,
"grad_norm": 0.22470302879810333,
"learning_rate": 0.003,
"loss": 3.0979190063476563,
"step": 120700
},
{
"epoch": 0.2638915816291701,
"grad_norm": 0.25222650170326233,
"learning_rate": 0.003,
"loss": 3.1603189086914063,
"step": 120800
},
{
"epoch": 0.2641100349252207,
"grad_norm": 0.15623459219932556,
"learning_rate": 0.003,
"loss": 3.146209411621094,
"step": 120900
},
{
"epoch": 0.26432848822127136,
"grad_norm": 0.19765418767929077,
"learning_rate": 0.003,
"loss": 3.118283996582031,
"step": 121000
},
{
"epoch": 0.264546941517322,
"grad_norm": 0.27228718996047974,
"learning_rate": 0.003,
"loss": 3.12953125,
"step": 121100
},
{
"epoch": 0.26476539481337263,
"grad_norm": 0.19979190826416016,
"learning_rate": 0.003,
"loss": 3.1466134643554686,
"step": 121200
},
{
"epoch": 0.26498384810942327,
"grad_norm": 0.18945246934890747,
"learning_rate": 0.003,
"loss": 3.1670947265625,
"step": 121300
},
{
"epoch": 0.2652023014054739,
"grad_norm": 0.18201249837875366,
"learning_rate": 0.003,
"loss": 3.1588299560546873,
"step": 121400
},
{
"epoch": 0.26542075470152454,
"grad_norm": 0.18221046030521393,
"learning_rate": 0.003,
"loss": 3.177774658203125,
"step": 121500
},
{
"epoch": 0.2656392079975752,
"grad_norm": 0.178348109126091,
"learning_rate": 0.003,
"loss": 3.1504315185546874,
"step": 121600
},
{
"epoch": 0.2658576612936258,
"grad_norm": 0.17613326013088226,
"learning_rate": 0.003,
"loss": 3.1706024169921876,
"step": 121700
},
{
"epoch": 0.26607611458967645,
"grad_norm": 0.1819799244403839,
"learning_rate": 0.003,
"loss": 3.1582550048828124,
"step": 121800
},
{
"epoch": 0.2662945678857271,
"grad_norm": 0.21169109642505646,
"learning_rate": 0.003,
"loss": 3.1287002563476562,
"step": 121900
},
{
"epoch": 0.2665130211817777,
"grad_norm": 0.17658387124538422,
"learning_rate": 0.003,
"loss": 3.1393206787109373,
"step": 122000
},
{
"epoch": 0.26673147447782836,
"grad_norm": 0.17100141942501068,
"learning_rate": 0.003,
"loss": 3.148631896972656,
"step": 122100
},
{
"epoch": 0.266949927773879,
"grad_norm": 0.1690492033958435,
"learning_rate": 0.003,
"loss": 3.1369088745117186,
"step": 122200
},
{
"epoch": 0.26716838106992963,
"grad_norm": 0.21133515238761902,
"learning_rate": 0.003,
"loss": 3.1234674072265625,
"step": 122300
},
{
"epoch": 0.26738683436598026,
"grad_norm": 0.1449619084596634,
"learning_rate": 0.003,
"loss": 3.2062924194335936,
"step": 122400
},
{
"epoch": 0.2676052876620309,
"grad_norm": 0.1649831086397171,
"learning_rate": 0.003,
"loss": 3.1165286254882814,
"step": 122500
},
{
"epoch": 0.26782374095808154,
"grad_norm": 0.14208738505840302,
"learning_rate": 0.003,
"loss": 3.1222979736328127,
"step": 122600
},
{
"epoch": 0.26804219425413217,
"grad_norm": 0.2803119421005249,
"learning_rate": 0.003,
"loss": 3.1634332275390626,
"step": 122700
},
{
"epoch": 0.2682606475501828,
"grad_norm": 0.1829257309436798,
"learning_rate": 0.003,
"loss": 3.1225985717773437,
"step": 122800
},
{
"epoch": 0.26847910084623344,
"grad_norm": 0.15668363869190216,
"learning_rate": 0.003,
"loss": 3.0825006103515626,
"step": 122900
},
{
"epoch": 0.2686975541422841,
"grad_norm": 0.18942736089229584,
"learning_rate": 0.003,
"loss": 3.1140615844726565,
"step": 123000
},
{
"epoch": 0.2689160074383347,
"grad_norm": 0.1751144975423813,
"learning_rate": 0.003,
"loss": 3.1174212646484376,
"step": 123100
},
{
"epoch": 0.26913446073438535,
"grad_norm": 0.15149031579494476,
"learning_rate": 0.003,
"loss": 3.092795715332031,
"step": 123200
},
{
"epoch": 0.269352914030436,
"grad_norm": 0.146224707365036,
"learning_rate": 0.003,
"loss": 3.1540908813476562,
"step": 123300
},
{
"epoch": 0.2695713673264866,
"grad_norm": 0.26576176285743713,
"learning_rate": 0.003,
"loss": 3.118391418457031,
"step": 123400
},
{
"epoch": 0.26978982062253726,
"grad_norm": 0.17119044065475464,
"learning_rate": 0.003,
"loss": 3.145680236816406,
"step": 123500
},
{
"epoch": 0.2700082739185879,
"grad_norm": 0.18358702957630157,
"learning_rate": 0.003,
"loss": 3.122941589355469,
"step": 123600
},
{
"epoch": 0.27022672721463853,
"grad_norm": 0.1615283489227295,
"learning_rate": 0.003,
"loss": 3.1004730224609376,
"step": 123700
},
{
"epoch": 0.2704451805106892,
"grad_norm": 0.16417482495307922,
"learning_rate": 0.003,
"loss": 3.1084674072265623,
"step": 123800
},
{
"epoch": 0.27066363380673986,
"grad_norm": 0.15363475680351257,
"learning_rate": 0.003,
"loss": 3.1380364990234373,
"step": 123900
},
{
"epoch": 0.2708820871027905,
"grad_norm": 0.21433115005493164,
"learning_rate": 0.003,
"loss": 3.09521728515625,
"step": 124000
},
{
"epoch": 0.2711005403988411,
"grad_norm": 0.16247905790805817,
"learning_rate": 0.003,
"loss": 2.980374755859375,
"step": 124100
},
{
"epoch": 0.27131899369489176,
"grad_norm": 0.23055775463581085,
"learning_rate": 0.003,
"loss": 3.0912255859375,
"step": 124200
},
{
"epoch": 0.2715374469909424,
"grad_norm": 0.17991212010383606,
"learning_rate": 0.003,
"loss": 3.075794677734375,
"step": 124300
},
{
"epoch": 0.27175590028699304,
"grad_norm": 0.15245744585990906,
"learning_rate": 0.003,
"loss": 3.1302386474609376,
"step": 124400
},
{
"epoch": 0.27197435358304367,
"grad_norm": 0.22990760207176208,
"learning_rate": 0.003,
"loss": 3.0966201782226563,
"step": 124500
},
{
"epoch": 0.2721928068790943,
"grad_norm": 0.14952386915683746,
"learning_rate": 0.003,
"loss": 3.107850341796875,
"step": 124600
},
{
"epoch": 0.27241126017514494,
"grad_norm": 0.1720409095287323,
"learning_rate": 0.003,
"loss": 3.147486572265625,
"step": 124700
},
{
"epoch": 0.2726297134711956,
"grad_norm": 0.16717100143432617,
"learning_rate": 0.003,
"loss": 3.1709051513671875,
"step": 124800
},
{
"epoch": 0.2728481667672462,
"grad_norm": 0.18736708164215088,
"learning_rate": 0.003,
"loss": 3.1149432373046877,
"step": 124900
},
{
"epoch": 0.27306662006329685,
"grad_norm": 0.1759713888168335,
"learning_rate": 0.003,
"loss": 3.0998104858398436,
"step": 125000
},
{
"epoch": 0.2732850733593475,
"grad_norm": 0.2139262855052948,
"learning_rate": 0.003,
"loss": 3.0760043334960936,
"step": 125100
},
{
"epoch": 0.2735035266553981,
"grad_norm": 0.1517336070537567,
"learning_rate": 0.003,
"loss": 3.1243695068359374,
"step": 125200
},
{
"epoch": 0.27372197995144876,
"grad_norm": 0.15442916750907898,
"learning_rate": 0.003,
"loss": 3.1157635498046874,
"step": 125300
},
{
"epoch": 0.2739404332474994,
"grad_norm": 0.1729786992073059,
"learning_rate": 0.003,
"loss": 3.130433349609375,
"step": 125400
},
{
"epoch": 0.27415888654355003,
"grad_norm": 0.172031968832016,
"learning_rate": 0.003,
"loss": 3.114808349609375,
"step": 125500
},
{
"epoch": 0.27437733983960066,
"grad_norm": 0.17771190404891968,
"learning_rate": 0.003,
"loss": 3.104170837402344,
"step": 125600
},
{
"epoch": 0.2745957931356513,
"grad_norm": 0.18168313801288605,
"learning_rate": 0.003,
"loss": 3.0751837158203124,
"step": 125700
},
{
"epoch": 0.27481424643170194,
"grad_norm": 0.16100353002548218,
"learning_rate": 0.003,
"loss": 3.1454086303710938,
"step": 125800
},
{
"epoch": 0.27503269972775257,
"grad_norm": 0.2005414068698883,
"learning_rate": 0.003,
"loss": 3.121745910644531,
"step": 125900
},
{
"epoch": 0.2752511530238032,
"grad_norm": 0.2184487283229828,
"learning_rate": 0.003,
"loss": 3.1026446533203127,
"step": 126000
},
{
"epoch": 0.27546960631985384,
"grad_norm": 0.17234332859516144,
"learning_rate": 0.003,
"loss": 3.1335687255859375,
"step": 126100
},
{
"epoch": 0.2756880596159045,
"grad_norm": 0.1576775461435318,
"learning_rate": 0.003,
"loss": 3.1129110717773436,
"step": 126200
},
{
"epoch": 0.2759065129119551,
"grad_norm": 0.14490582048892975,
"learning_rate": 0.003,
"loss": 3.134710693359375,
"step": 126300
},
{
"epoch": 0.27612496620800575,
"grad_norm": 0.23765777051448822,
"learning_rate": 0.003,
"loss": 3.111960754394531,
"step": 126400
},
{
"epoch": 0.2763434195040564,
"grad_norm": 0.19747157394886017,
"learning_rate": 0.003,
"loss": 3.0904254150390624,
"step": 126500
},
{
"epoch": 0.276561872800107,
"grad_norm": 0.2658751308917999,
"learning_rate": 0.003,
"loss": 3.0246206665039064,
"step": 126600
},
{
"epoch": 0.27678032609615766,
"grad_norm": 0.1605808287858963,
"learning_rate": 0.003,
"loss": 3.0998355102539064,
"step": 126700
},
{
"epoch": 0.2769987793922083,
"grad_norm": 0.165673166513443,
"learning_rate": 0.003,
"loss": 3.0695989990234374,
"step": 126800
},
{
"epoch": 0.27721723268825893,
"grad_norm": 0.18814611434936523,
"learning_rate": 0.003,
"loss": 3.0552288818359377,
"step": 126900
},
{
"epoch": 0.27743568598430957,
"grad_norm": 0.1825118511915207,
"learning_rate": 0.003,
"loss": 3.0807373046875,
"step": 127000
},
{
"epoch": 0.27765413928036026,
"grad_norm": 0.17635834217071533,
"learning_rate": 0.003,
"loss": 3.0948751831054686,
"step": 127100
},
{
"epoch": 0.2778725925764109,
"grad_norm": 0.2630055546760559,
"learning_rate": 0.003,
"loss": 3.0975680541992188,
"step": 127200
},
{
"epoch": 0.27809104587246153,
"grad_norm": 0.24227474629878998,
"learning_rate": 0.003,
"loss": 3.055248107910156,
"step": 127300
},
{
"epoch": 0.27830949916851216,
"grad_norm": 0.21120278537273407,
"learning_rate": 0.003,
"loss": 3.068297119140625,
"step": 127400
},
{
"epoch": 0.2785279524645628,
"grad_norm": 0.3816051483154297,
"learning_rate": 0.003,
"loss": 3.0776156616210937,
"step": 127500
},
{
"epoch": 0.27874640576061344,
"grad_norm": 0.16917425394058228,
"learning_rate": 0.003,
"loss": 3.0876223754882814,
"step": 127600
},
{
"epoch": 0.27896485905666407,
"grad_norm": 0.20668604969978333,
"learning_rate": 0.003,
"loss": 3.204374084472656,
"step": 127700
},
{
"epoch": 0.2791833123527147,
"grad_norm": 0.18532617390155792,
"learning_rate": 0.003,
"loss": 3.22772216796875,
"step": 127800
},
{
"epoch": 0.27940176564876534,
"grad_norm": 0.1958247721195221,
"learning_rate": 0.003,
"loss": 3.193952941894531,
"step": 127900
},
{
"epoch": 0.279620218944816,
"grad_norm": 0.22922642529010773,
"learning_rate": 0.003,
"loss": 3.187032470703125,
"step": 128000
},
{
"epoch": 0.2798386722408666,
"grad_norm": 0.24574919044971466,
"learning_rate": 0.003,
"loss": 3.2100088500976565,
"step": 128100
},
{
"epoch": 0.28005712553691725,
"grad_norm": 0.1636534333229065,
"learning_rate": 0.003,
"loss": 3.194091796875,
"step": 128200
},
{
"epoch": 0.2802755788329679,
"grad_norm": 0.23779068887233734,
"learning_rate": 0.003,
"loss": 3.189336853027344,
"step": 128300
},
{
"epoch": 0.2804940321290185,
"grad_norm": 0.16079804301261902,
"learning_rate": 0.003,
"loss": 3.1519854736328123,
"step": 128400
},
{
"epoch": 0.28071248542506916,
"grad_norm": 0.1529242843389511,
"learning_rate": 0.003,
"loss": 3.179140625,
"step": 128500
},
{
"epoch": 0.2809309387211198,
"grad_norm": 0.24319961667060852,
"learning_rate": 0.003,
"loss": 3.177486572265625,
"step": 128600
},
{
"epoch": 0.28114939201717043,
"grad_norm": 0.18524886667728424,
"learning_rate": 0.003,
"loss": 3.190474548339844,
"step": 128700
},
{
"epoch": 0.28136784531322107,
"grad_norm": 0.19689129292964935,
"learning_rate": 0.003,
"loss": 3.174749450683594,
"step": 128800
},
{
"epoch": 0.2815862986092717,
"grad_norm": 0.21814829111099243,
"learning_rate": 0.003,
"loss": 3.2250634765625,
"step": 128900
},
{
"epoch": 0.28180475190532234,
"grad_norm": 0.27080363035202026,
"learning_rate": 0.003,
"loss": 3.1633697509765626,
"step": 129000
},
{
"epoch": 0.282023205201373,
"grad_norm": 0.18583400547504425,
"learning_rate": 0.003,
"loss": 3.1680859375,
"step": 129100
},
{
"epoch": 0.2822416584974236,
"grad_norm": 0.3391522467136383,
"learning_rate": 0.003,
"loss": 3.1928253173828125,
"step": 129200
},
{
"epoch": 0.28246011179347424,
"grad_norm": 0.23984627425670624,
"learning_rate": 0.003,
"loss": 3.1604623413085937,
"step": 129300
},
{
"epoch": 0.2826785650895249,
"grad_norm": 0.19537843763828278,
"learning_rate": 0.003,
"loss": 3.2140252685546873,
"step": 129400
},
{
"epoch": 0.2828970183855755,
"grad_norm": 0.16202545166015625,
"learning_rate": 0.003,
"loss": 3.1751544189453127,
"step": 129500
},
{
"epoch": 0.28311547168162615,
"grad_norm": 0.15761688351631165,
"learning_rate": 0.003,
"loss": 3.1904876708984373,
"step": 129600
},
{
"epoch": 0.2833339249776768,
"grad_norm": 0.24658435583114624,
"learning_rate": 0.003,
"loss": 3.164392395019531,
"step": 129700
},
{
"epoch": 0.2835523782737274,
"grad_norm": 0.18049539625644684,
"learning_rate": 0.003,
"loss": 3.2037832641601565,
"step": 129800
},
{
"epoch": 0.28377083156977806,
"grad_norm": 0.23034675419330597,
"learning_rate": 0.003,
"loss": 3.2158367919921873,
"step": 129900
},
{
"epoch": 0.2839892848658287,
"grad_norm": 0.2013300359249115,
"learning_rate": 0.003,
"loss": 3.226523132324219,
"step": 130000
},
{
"epoch": 0.28420773816187933,
"grad_norm": 0.1996563971042633,
"learning_rate": 0.003,
"loss": 3.176661376953125,
"step": 130100
},
{
"epoch": 0.28442619145792997,
"grad_norm": 0.23476724326610565,
"learning_rate": 0.003,
"loss": 3.1794467163085938,
"step": 130200
},
{
"epoch": 0.2846446447539806,
"grad_norm": 0.16203320026397705,
"learning_rate": 0.003,
"loss": 3.1501620483398436,
"step": 130300
},
{
"epoch": 0.2848630980500313,
"grad_norm": 0.22591614723205566,
"learning_rate": 0.003,
"loss": 3.1584625244140625,
"step": 130400
},
{
"epoch": 0.28508155134608193,
"grad_norm": 0.15424208343029022,
"learning_rate": 0.003,
"loss": 3.18368896484375,
"step": 130500
},
{
"epoch": 0.28530000464213257,
"grad_norm": 0.1743527203798294,
"learning_rate": 0.003,
"loss": 3.1977545166015626,
"step": 130600
},
{
"epoch": 0.2855184579381832,
"grad_norm": 0.15016095340251923,
"learning_rate": 0.003,
"loss": 3.175804443359375,
"step": 130700
},
{
"epoch": 0.28573691123423384,
"grad_norm": 0.18600404262542725,
"learning_rate": 0.003,
"loss": 3.1796951293945312,
"step": 130800
},
{
"epoch": 0.2859553645302845,
"grad_norm": 0.19670964777469635,
"learning_rate": 0.003,
"loss": 3.1544488525390624,
"step": 130900
},
{
"epoch": 0.2861738178263351,
"grad_norm": 0.1564444750547409,
"learning_rate": 0.003,
"loss": 3.13692138671875,
"step": 131000
},
{
"epoch": 0.28639227112238574,
"grad_norm": 0.1812867373228073,
"learning_rate": 0.003,
"loss": 3.164166259765625,
"step": 131100
},
{
"epoch": 0.2866107244184364,
"grad_norm": 0.16804924607276917,
"learning_rate": 0.003,
"loss": 3.13121337890625,
"step": 131200
},
{
"epoch": 0.286829177714487,
"grad_norm": 0.1703408658504486,
"learning_rate": 0.003,
"loss": 3.2148712158203123,
"step": 131300
},
{
"epoch": 0.28704763101053765,
"grad_norm": 0.2654840052127838,
"learning_rate": 0.003,
"loss": 3.1748138427734376,
"step": 131400
},
{
"epoch": 0.2872660843065883,
"grad_norm": 0.14809302985668182,
"learning_rate": 0.003,
"loss": 3.2007412719726562,
"step": 131500
},
{
"epoch": 0.2874845376026389,
"grad_norm": 0.2317712903022766,
"learning_rate": 0.003,
"loss": 3.1763409423828124,
"step": 131600
},
{
"epoch": 0.28770299089868956,
"grad_norm": 0.17288662493228912,
"learning_rate": 0.003,
"loss": 3.228929443359375,
"step": 131700
},
{
"epoch": 0.2879214441947402,
"grad_norm": 0.15619529783725739,
"learning_rate": 0.003,
"loss": 3.1699227905273437,
"step": 131800
},
{
"epoch": 0.28813989749079083,
"grad_norm": 0.20998290181159973,
"learning_rate": 0.003,
"loss": 3.1900537109375,
"step": 131900
},
{
"epoch": 0.28835835078684147,
"grad_norm": 0.16902107000350952,
"learning_rate": 0.003,
"loss": 3.1908218383789064,
"step": 132000
},
{
"epoch": 0.2885768040828921,
"grad_norm": 0.17402586340904236,
"learning_rate": 0.003,
"loss": 3.1593307495117187,
"step": 132100
},
{
"epoch": 0.28879525737894274,
"grad_norm": 0.2578791081905365,
"learning_rate": 0.003,
"loss": 3.1821502685546874,
"step": 132200
},
{
"epoch": 0.2890137106749934,
"grad_norm": 0.1539982259273529,
"learning_rate": 0.003,
"loss": 3.24119140625,
"step": 132300
},
{
"epoch": 0.289232163971044,
"grad_norm": 0.17721042037010193,
"learning_rate": 0.003,
"loss": 3.242032470703125,
"step": 132400
},
{
"epoch": 0.28945061726709465,
"grad_norm": 0.16132225096225739,
"learning_rate": 0.003,
"loss": 3.213353576660156,
"step": 132500
},
{
"epoch": 0.2896690705631453,
"grad_norm": 0.16307580471038818,
"learning_rate": 0.003,
"loss": 3.2237362670898437,
"step": 132600
},
{
"epoch": 0.2898875238591959,
"grad_norm": 0.19305144250392914,
"learning_rate": 0.003,
"loss": 3.1388040161132813,
"step": 132700
},
{
"epoch": 0.29010597715524655,
"grad_norm": 0.19119708240032196,
"learning_rate": 0.003,
"loss": 3.195961608886719,
"step": 132800
},
{
"epoch": 0.2903244304512972,
"grad_norm": 0.20485588908195496,
"learning_rate": 0.003,
"loss": 3.1509246826171875,
"step": 132900
},
{
"epoch": 0.2905428837473478,
"grad_norm": 0.21030651032924652,
"learning_rate": 0.003,
"loss": 3.201534423828125,
"step": 133000
},
{
"epoch": 0.29076133704339846,
"grad_norm": 0.17510755360126495,
"learning_rate": 0.003,
"loss": 3.199417724609375,
"step": 133100
},
{
"epoch": 0.2909797903394491,
"grad_norm": 0.32947033643722534,
"learning_rate": 0.003,
"loss": 3.2026324462890625,
"step": 133200
},
{
"epoch": 0.29119824363549973,
"grad_norm": 0.20043914020061493,
"learning_rate": 0.003,
"loss": 3.182317199707031,
"step": 133300
},
{
"epoch": 0.29141669693155037,
"grad_norm": 0.19705235958099365,
"learning_rate": 0.003,
"loss": 3.2217138671875,
"step": 133400
},
{
"epoch": 0.291635150227601,
"grad_norm": 0.16933363676071167,
"learning_rate": 0.003,
"loss": 3.2272128295898437,
"step": 133500
},
{
"epoch": 0.29185360352365164,
"grad_norm": 0.16237424314022064,
"learning_rate": 0.003,
"loss": 3.19873291015625,
"step": 133600
},
{
"epoch": 0.29207205681970233,
"grad_norm": 0.1812378317117691,
"learning_rate": 0.003,
"loss": 3.180633850097656,
"step": 133700
},
{
"epoch": 0.29229051011575297,
"grad_norm": 0.17355945706367493,
"learning_rate": 0.003,
"loss": 3.2264837646484374,
"step": 133800
},
{
"epoch": 0.2925089634118036,
"grad_norm": 0.17370428144931793,
"learning_rate": 0.003,
"loss": 3.176700439453125,
"step": 133900
},
{
"epoch": 0.29272741670785424,
"grad_norm": 0.24076209962368011,
"learning_rate": 0.003,
"loss": 3.1999746704101564,
"step": 134000
},
{
"epoch": 0.2929458700039049,
"grad_norm": 0.1681448370218277,
"learning_rate": 0.003,
"loss": 3.1995059204101564,
"step": 134100
},
{
"epoch": 0.2931643232999555,
"grad_norm": 0.2963580787181854,
"learning_rate": 0.003,
"loss": 3.2250381469726563,
"step": 134200
},
{
"epoch": 0.29338277659600615,
"grad_norm": 0.16434602439403534,
"learning_rate": 0.003,
"loss": 3.1836260986328124,
"step": 134300
},
{
"epoch": 0.2936012298920568,
"grad_norm": 0.19248543679714203,
"learning_rate": 0.003,
"loss": 3.2431411743164062,
"step": 134400
},
{
"epoch": 0.2938196831881074,
"grad_norm": 0.2197428047657013,
"learning_rate": 0.003,
"loss": 3.1926956176757812,
"step": 134500
},
{
"epoch": 0.29403813648415805,
"grad_norm": 0.1991998553276062,
"learning_rate": 0.003,
"loss": 3.1818472290039064,
"step": 134600
},
{
"epoch": 0.2942565897802087,
"grad_norm": 0.18852558732032776,
"learning_rate": 0.003,
"loss": 3.1907608032226564,
"step": 134700
},
{
"epoch": 0.2944750430762593,
"grad_norm": 0.20869626104831696,
"learning_rate": 0.003,
"loss": 3.1867388916015624,
"step": 134800
},
{
"epoch": 0.29469349637230996,
"grad_norm": 0.17816461622714996,
"learning_rate": 0.003,
"loss": 3.131686096191406,
"step": 134900
},
{
"epoch": 0.2949119496683606,
"grad_norm": 0.1907869577407837,
"learning_rate": 0.003,
"loss": 3.214499816894531,
"step": 135000
},
{
"epoch": 0.29513040296441123,
"grad_norm": 0.15174990892410278,
"learning_rate": 0.003,
"loss": 3.210446472167969,
"step": 135100
},
{
"epoch": 0.29534885626046187,
"grad_norm": 0.20105698704719543,
"learning_rate": 0.003,
"loss": 3.1645794677734376,
"step": 135200
},
{
"epoch": 0.2955673095565125,
"grad_norm": 0.19103233516216278,
"learning_rate": 0.003,
"loss": 3.164373474121094,
"step": 135300
},
{
"epoch": 0.29578576285256314,
"grad_norm": 0.1916809380054474,
"learning_rate": 0.003,
"loss": 3.1695938110351562,
"step": 135400
},
{
"epoch": 0.2960042161486138,
"grad_norm": 0.1843739002943039,
"learning_rate": 0.003,
"loss": 3.1917315673828126,
"step": 135500
},
{
"epoch": 0.2962226694446644,
"grad_norm": 0.16667531430721283,
"learning_rate": 0.003,
"loss": 3.1886322021484377,
"step": 135600
},
{
"epoch": 0.29644112274071505,
"grad_norm": 0.21029360592365265,
"learning_rate": 0.003,
"loss": 3.210748596191406,
"step": 135700
},
{
"epoch": 0.2966595760367657,
"grad_norm": 0.17889174818992615,
"learning_rate": 0.003,
"loss": 3.1713192749023436,
"step": 135800
},
{
"epoch": 0.2968780293328163,
"grad_norm": 0.17533376812934875,
"learning_rate": 0.003,
"loss": 3.2180303955078124,
"step": 135900
},
{
"epoch": 0.29709648262886695,
"grad_norm": 0.1637962907552719,
"learning_rate": 0.003,
"loss": 3.2227993774414063,
"step": 136000
},
{
"epoch": 0.2973149359249176,
"grad_norm": 0.3537386953830719,
"learning_rate": 0.003,
"loss": 3.2023394775390623,
"step": 136100
},
{
"epoch": 0.2975333892209682,
"grad_norm": 0.19465351104736328,
"learning_rate": 0.003,
"loss": 3.2192376708984374,
"step": 136200
},
{
"epoch": 0.29775184251701886,
"grad_norm": 0.2649924159049988,
"learning_rate": 0.003,
"loss": 3.177152099609375,
"step": 136300
},
{
"epoch": 0.2979702958130695,
"grad_norm": 0.1992715746164322,
"learning_rate": 0.003,
"loss": 3.209580078125,
"step": 136400
},
{
"epoch": 0.29818874910912013,
"grad_norm": 0.16940732300281525,
"learning_rate": 0.003,
"loss": 3.2093356323242186,
"step": 136500
},
{
"epoch": 0.29840720240517077,
"grad_norm": 0.16772933304309845,
"learning_rate": 0.003,
"loss": 3.221116638183594,
"step": 136600
},
{
"epoch": 0.2986256557012214,
"grad_norm": 0.1964656412601471,
"learning_rate": 0.003,
"loss": 3.2100775146484377,
"step": 136700
},
{
"epoch": 0.29884410899727204,
"grad_norm": 0.2089720517396927,
"learning_rate": 0.003,
"loss": 3.203949279785156,
"step": 136800
},
{
"epoch": 0.2990625622933227,
"grad_norm": 0.15475992858409882,
"learning_rate": 0.003,
"loss": 3.205782470703125,
"step": 136900
},
{
"epoch": 0.29928101558937337,
"grad_norm": 0.1778334528207779,
"learning_rate": 0.003,
"loss": 3.2443170166015625,
"step": 137000
},
{
"epoch": 0.299499468885424,
"grad_norm": 0.1804296225309372,
"learning_rate": 0.003,
"loss": 3.21103515625,
"step": 137100
},
{
"epoch": 0.29971792218147464,
"grad_norm": 0.20873281359672546,
"learning_rate": 0.003,
"loss": 3.2040234375,
"step": 137200
},
{
"epoch": 0.2999363754775253,
"grad_norm": 0.24339774250984192,
"learning_rate": 0.003,
"loss": 3.1806402587890625,
"step": 137300
},
{
"epoch": 0.3001548287735759,
"grad_norm": 0.22704702615737915,
"learning_rate": 0.003,
"loss": 3.1972055053710937,
"step": 137400
},
{
"epoch": 0.30037328206962655,
"grad_norm": 0.15578119456768036,
"learning_rate": 0.003,
"loss": 3.1968524169921877,
"step": 137500
},
{
"epoch": 0.3005917353656772,
"grad_norm": 0.2116861343383789,
"learning_rate": 0.003,
"loss": 3.20760009765625,
"step": 137600
},
{
"epoch": 0.3008101886617278,
"grad_norm": 0.15658578276634216,
"learning_rate": 0.003,
"loss": 3.220193176269531,
"step": 137700
},
{
"epoch": 0.30102864195777845,
"grad_norm": 0.18922600150108337,
"learning_rate": 0.003,
"loss": 3.1731051635742187,
"step": 137800
},
{
"epoch": 0.3012470952538291,
"grad_norm": 0.16404539346694946,
"learning_rate": 0.003,
"loss": 3.1814349365234373,
"step": 137900
},
{
"epoch": 0.3014655485498797,
"grad_norm": 0.20131675899028778,
"learning_rate": 0.003,
"loss": 3.1948602294921873,
"step": 138000
},
{
"epoch": 0.30168400184593036,
"grad_norm": 0.17711040377616882,
"learning_rate": 0.003,
"loss": 3.2230322265625,
"step": 138100
},
{
"epoch": 0.301902455141981,
"grad_norm": 0.1827816516160965,
"learning_rate": 0.003,
"loss": 3.1611846923828124,
"step": 138200
},
{
"epoch": 0.30212090843803163,
"grad_norm": 0.23586484789848328,
"learning_rate": 0.003,
"loss": 3.201513671875,
"step": 138300
},
{
"epoch": 0.30233936173408227,
"grad_norm": 0.20572659373283386,
"learning_rate": 0.003,
"loss": 3.195331726074219,
"step": 138400
},
{
"epoch": 0.3025578150301329,
"grad_norm": 0.18462501466274261,
"learning_rate": 0.003,
"loss": 3.206833801269531,
"step": 138500
},
{
"epoch": 0.30277626832618354,
"grad_norm": 0.17236556112766266,
"learning_rate": 0.003,
"loss": 3.178404541015625,
"step": 138600
},
{
"epoch": 0.3029947216222342,
"grad_norm": 0.15189675986766815,
"learning_rate": 0.003,
"loss": 3.1943118286132814,
"step": 138700
},
{
"epoch": 0.3032131749182848,
"grad_norm": 0.1795898973941803,
"learning_rate": 0.003,
"loss": 3.2082452392578125,
"step": 138800
},
{
"epoch": 0.30343162821433545,
"grad_norm": 0.24792973697185516,
"learning_rate": 0.003,
"loss": 3.2121905517578124,
"step": 138900
},
{
"epoch": 0.3036500815103861,
"grad_norm": 0.1631484478712082,
"learning_rate": 0.003,
"loss": 3.1948968505859376,
"step": 139000
},
{
"epoch": 0.3038685348064367,
"grad_norm": 0.17224180698394775,
"learning_rate": 0.003,
"loss": 3.2479281616210938,
"step": 139100
},
{
"epoch": 0.30408698810248735,
"grad_norm": 0.18179869651794434,
"learning_rate": 0.003,
"loss": 3.195875549316406,
"step": 139200
},
{
"epoch": 0.304305441398538,
"grad_norm": 0.25769829750061035,
"learning_rate": 0.003,
"loss": 3.20810302734375,
"step": 139300
},
{
"epoch": 0.3045238946945886,
"grad_norm": 0.19908572733402252,
"learning_rate": 0.003,
"loss": 3.1936154174804687,
"step": 139400
},
{
"epoch": 0.30474234799063926,
"grad_norm": 0.15726956725120544,
"learning_rate": 0.003,
"loss": 3.17328125,
"step": 139500
},
{
"epoch": 0.3049608012866899,
"grad_norm": 0.22366781532764435,
"learning_rate": 0.003,
"loss": 3.179468078613281,
"step": 139600
},
{
"epoch": 0.30517925458274053,
"grad_norm": 0.1611061990261078,
"learning_rate": 0.003,
"loss": 3.19077880859375,
"step": 139700
},
{
"epoch": 0.30539770787879117,
"grad_norm": 0.17893096804618835,
"learning_rate": 0.003,
"loss": 3.221007995605469,
"step": 139800
},
{
"epoch": 0.3056161611748418,
"grad_norm": 0.19288408756256104,
"learning_rate": 0.003,
"loss": 3.218514709472656,
"step": 139900
},
{
"epoch": 0.30583461447089244,
"grad_norm": 0.2554208040237427,
"learning_rate": 0.003,
"loss": 2.9139288330078124,
"step": 140000
},
{
"epoch": 0.3060530677669431,
"grad_norm": 0.1867968589067459,
"learning_rate": 0.003,
"loss": 3.2304754638671875,
"step": 140100
},
{
"epoch": 0.3062715210629937,
"grad_norm": 0.19115705788135529,
"learning_rate": 0.003,
"loss": 3.1553607177734375,
"step": 140200
},
{
"epoch": 0.3064899743590444,
"grad_norm": 0.1747879832983017,
"learning_rate": 0.003,
"loss": 3.248841552734375,
"step": 140300
},
{
"epoch": 0.30670842765509504,
"grad_norm": 0.21023240685462952,
"learning_rate": 0.003,
"loss": 3.2210183715820313,
"step": 140400
},
{
"epoch": 0.3069268809511457,
"grad_norm": 0.1795470267534256,
"learning_rate": 0.003,
"loss": 3.18628662109375,
"step": 140500
},
{
"epoch": 0.3071453342471963,
"grad_norm": 0.17351393401622772,
"learning_rate": 0.003,
"loss": 3.19286865234375,
"step": 140600
},
{
"epoch": 0.30736378754324695,
"grad_norm": 0.18436656892299652,
"learning_rate": 0.003,
"loss": 3.205523681640625,
"step": 140700
},
{
"epoch": 0.3075822408392976,
"grad_norm": 0.15725570917129517,
"learning_rate": 0.003,
"loss": 3.230887145996094,
"step": 140800
},
{
"epoch": 0.3078006941353482,
"grad_norm": 0.21773605048656464,
"learning_rate": 0.003,
"loss": 3.228748779296875,
"step": 140900
},
{
"epoch": 0.30801914743139885,
"grad_norm": 0.17027831077575684,
"learning_rate": 0.003,
"loss": 3.186734313964844,
"step": 141000
},
{
"epoch": 0.3082376007274495,
"grad_norm": 0.15384380519390106,
"learning_rate": 0.003,
"loss": 3.1859201049804686,
"step": 141100
},
{
"epoch": 0.3084560540235001,
"grad_norm": 0.3453856110572815,
"learning_rate": 0.003,
"loss": 3.160126037597656,
"step": 141200
},
{
"epoch": 0.30867450731955076,
"grad_norm": 0.327231764793396,
"learning_rate": 0.003,
"loss": 3.1771490478515627,
"step": 141300
},
{
"epoch": 0.3088929606156014,
"grad_norm": 0.28434672951698303,
"learning_rate": 0.003,
"loss": 3.1603729248046877,
"step": 141400
},
{
"epoch": 0.30911141391165203,
"grad_norm": 0.20859956741333008,
"learning_rate": 0.003,
"loss": 3.199524230957031,
"step": 141500
},
{
"epoch": 0.30932986720770267,
"grad_norm": 0.16180141270160675,
"learning_rate": 0.003,
"loss": 3.217564697265625,
"step": 141600
},
{
"epoch": 0.3095483205037533,
"grad_norm": 0.18180030584335327,
"learning_rate": 0.003,
"loss": 3.2199407958984376,
"step": 141700
},
{
"epoch": 0.30976677379980394,
"grad_norm": 0.16201762855052948,
"learning_rate": 0.003,
"loss": 3.174957275390625,
"step": 141800
},
{
"epoch": 0.3099852270958546,
"grad_norm": 0.1949051469564438,
"learning_rate": 0.003,
"loss": 3.217715148925781,
"step": 141900
},
{
"epoch": 0.3102036803919052,
"grad_norm": 0.15087030827999115,
"learning_rate": 0.003,
"loss": 3.186820068359375,
"step": 142000
},
{
"epoch": 0.31042213368795585,
"grad_norm": 0.168676495552063,
"learning_rate": 0.003,
"loss": 3.1756060791015623,
"step": 142100
},
{
"epoch": 0.3106405869840065,
"grad_norm": 0.24860401451587677,
"learning_rate": 0.003,
"loss": 3.2399252319335936,
"step": 142200
},
{
"epoch": 0.3108590402800571,
"grad_norm": 0.15668369829654694,
"learning_rate": 0.003,
"loss": 3.199765319824219,
"step": 142300
},
{
"epoch": 0.31107749357610776,
"grad_norm": 0.6814060807228088,
"learning_rate": 0.003,
"loss": 3.2036663818359377,
"step": 142400
},
{
"epoch": 0.3112959468721584,
"grad_norm": 0.1491079330444336,
"learning_rate": 0.003,
"loss": 3.225286560058594,
"step": 142500
},
{
"epoch": 0.311514400168209,
"grad_norm": 0.15027421712875366,
"learning_rate": 0.003,
"loss": 3.2134439086914064,
"step": 142600
},
{
"epoch": 0.31173285346425966,
"grad_norm": 0.2214614301919937,
"learning_rate": 0.003,
"loss": 3.1999752807617186,
"step": 142700
},
{
"epoch": 0.3119513067603103,
"grad_norm": 0.18709635734558105,
"learning_rate": 0.003,
"loss": 3.1740081787109373,
"step": 142800
},
{
"epoch": 0.31216976005636093,
"grad_norm": 0.24174664914608002,
"learning_rate": 0.003,
"loss": 3.2188986206054686,
"step": 142900
},
{
"epoch": 0.31238821335241157,
"grad_norm": 0.14840149879455566,
"learning_rate": 0.003,
"loss": 3.203288879394531,
"step": 143000
},
{
"epoch": 0.3126066666484622,
"grad_norm": 0.3073531687259674,
"learning_rate": 0.003,
"loss": 3.257579650878906,
"step": 143100
},
{
"epoch": 0.31282511994451284,
"grad_norm": 0.21026237308979034,
"learning_rate": 0.003,
"loss": 3.2264254760742186,
"step": 143200
},
{
"epoch": 0.3130435732405635,
"grad_norm": 0.16831256449222565,
"learning_rate": 0.003,
"loss": 3.1818194580078125,
"step": 143300
},
{
"epoch": 0.3132620265366141,
"grad_norm": 0.16093488037586212,
"learning_rate": 0.003,
"loss": 3.1373519897460938,
"step": 143400
},
{
"epoch": 0.31348047983266475,
"grad_norm": 0.18346524238586426,
"learning_rate": 0.003,
"loss": 3.1911004638671874,
"step": 143500
},
{
"epoch": 0.31369893312871544,
"grad_norm": 0.18986782431602478,
"learning_rate": 0.003,
"loss": 3.177647705078125,
"step": 143600
},
{
"epoch": 0.3139173864247661,
"grad_norm": 0.22777634859085083,
"learning_rate": 0.003,
"loss": 3.1712738037109376,
"step": 143700
},
{
"epoch": 0.3141358397208167,
"grad_norm": 0.14152170717716217,
"learning_rate": 0.003,
"loss": 3.2470083618164063,
"step": 143800
},
{
"epoch": 0.31435429301686735,
"grad_norm": 0.21601779758930206,
"learning_rate": 0.003,
"loss": 3.2211138916015627,
"step": 143900
},
{
"epoch": 0.314572746312918,
"grad_norm": 0.16085198521614075,
"learning_rate": 0.003,
"loss": 3.1816421508789063,
"step": 144000
},
{
"epoch": 0.3147911996089686,
"grad_norm": 0.1805422604084015,
"learning_rate": 0.003,
"loss": 3.199898376464844,
"step": 144100
},
{
"epoch": 0.31500965290501926,
"grad_norm": 0.2308289259672165,
"learning_rate": 0.003,
"loss": 3.208460388183594,
"step": 144200
},
{
"epoch": 0.3152281062010699,
"grad_norm": 0.17883312702178955,
"learning_rate": 0.003,
"loss": 3.149988098144531,
"step": 144300
},
{
"epoch": 0.3154465594971205,
"grad_norm": 0.17172366380691528,
"learning_rate": 0.003,
"loss": 3.211492919921875,
"step": 144400
},
{
"epoch": 0.31566501279317116,
"grad_norm": 0.3454696536064148,
"learning_rate": 0.003,
"loss": 3.1982745361328124,
"step": 144500
},
{
"epoch": 0.3158834660892218,
"grad_norm": 0.18591400980949402,
"learning_rate": 0.003,
"loss": 3.2074407958984374,
"step": 144600
},
{
"epoch": 0.31610191938527243,
"grad_norm": 0.1585206687450409,
"learning_rate": 0.003,
"loss": 3.1934228515625,
"step": 144700
},
{
"epoch": 0.31632037268132307,
"grad_norm": 0.34315547347068787,
"learning_rate": 0.003,
"loss": 3.1926708984375,
"step": 144800
},
{
"epoch": 0.3165388259773737,
"grad_norm": 0.17863717675209045,
"learning_rate": 0.003,
"loss": 3.1578744506835936,
"step": 144900
},
{
"epoch": 0.31675727927342434,
"grad_norm": 0.24649377167224884,
"learning_rate": 0.003,
"loss": 3.197144775390625,
"step": 145000
},
{
"epoch": 0.316975732569475,
"grad_norm": 0.16562138497829437,
"learning_rate": 0.003,
"loss": 3.191086730957031,
"step": 145100
},
{
"epoch": 0.3171941858655256,
"grad_norm": 0.17314296960830688,
"learning_rate": 0.003,
"loss": 3.2123272705078123,
"step": 145200
},
{
"epoch": 0.31741263916157625,
"grad_norm": 0.16946852207183838,
"learning_rate": 0.003,
"loss": 3.179129943847656,
"step": 145300
},
{
"epoch": 0.3176310924576269,
"grad_norm": 0.17480599880218506,
"learning_rate": 0.003,
"loss": 3.2248297119140625,
"step": 145400
},
{
"epoch": 0.3178495457536775,
"grad_norm": 0.16407208144664764,
"learning_rate": 0.003,
"loss": 3.196800537109375,
"step": 145500
},
{
"epoch": 0.31806799904972816,
"grad_norm": 0.20925581455230713,
"learning_rate": 0.003,
"loss": 3.18634521484375,
"step": 145600
},
{
"epoch": 0.3182864523457788,
"grad_norm": 0.222587451338768,
"learning_rate": 0.003,
"loss": 3.2120587158203127,
"step": 145700
},
{
"epoch": 0.31850490564182943,
"grad_norm": 0.17559437453746796,
"learning_rate": 0.003,
"loss": 3.2094912719726563,
"step": 145800
},
{
"epoch": 0.31872335893788006,
"grad_norm": 0.253233939409256,
"learning_rate": 0.003,
"loss": 3.1960491943359375,
"step": 145900
},
{
"epoch": 0.3189418122339307,
"grad_norm": 0.15889482200145721,
"learning_rate": 0.003,
"loss": 3.218516845703125,
"step": 146000
},
{
"epoch": 0.31916026552998134,
"grad_norm": 0.23230864107608795,
"learning_rate": 0.003,
"loss": 3.1856573486328124,
"step": 146100
},
{
"epoch": 0.31937871882603197,
"grad_norm": 0.14948810636997223,
"learning_rate": 0.003,
"loss": 3.2190350341796874,
"step": 146200
},
{
"epoch": 0.3195971721220826,
"grad_norm": 0.5103902816772461,
"learning_rate": 0.003,
"loss": 3.1645404052734376,
"step": 146300
},
{
"epoch": 0.31981562541813324,
"grad_norm": 0.1744993031024933,
"learning_rate": 0.003,
"loss": 3.2244082641601564,
"step": 146400
},
{
"epoch": 0.3200340787141839,
"grad_norm": 0.17527836561203003,
"learning_rate": 0.003,
"loss": 3.2356802368164064,
"step": 146500
},
{
"epoch": 0.3202525320102345,
"grad_norm": 0.30127814412117004,
"learning_rate": 0.003,
"loss": 3.2353482055664062,
"step": 146600
},
{
"epoch": 0.32047098530628515,
"grad_norm": 0.1600414365530014,
"learning_rate": 0.003,
"loss": 3.21412353515625,
"step": 146700
},
{
"epoch": 0.3206894386023358,
"grad_norm": 0.17066684365272522,
"learning_rate": 0.003,
"loss": 3.1851055908203123,
"step": 146800
},
{
"epoch": 0.3209078918983865,
"grad_norm": 0.17907121777534485,
"learning_rate": 0.003,
"loss": 3.18966552734375,
"step": 146900
},
{
"epoch": 0.3211263451944371,
"grad_norm": 0.16637155413627625,
"learning_rate": 0.003,
"loss": 3.190469055175781,
"step": 147000
},
{
"epoch": 0.32134479849048775,
"grad_norm": 0.19598500430583954,
"learning_rate": 0.003,
"loss": 3.1495361328125,
"step": 147100
},
{
"epoch": 0.3215632517865384,
"grad_norm": 0.17670966684818268,
"learning_rate": 0.003,
"loss": 3.193385009765625,
"step": 147200
},
{
"epoch": 0.321781705082589,
"grad_norm": 0.15445983409881592,
"learning_rate": 0.003,
"loss": 3.1991397094726564,
"step": 147300
},
{
"epoch": 0.32200015837863966,
"grad_norm": 0.18516457080841064,
"learning_rate": 0.003,
"loss": 3.094637145996094,
"step": 147400
},
{
"epoch": 0.3222186116746903,
"grad_norm": 0.19677746295928955,
"learning_rate": 0.003,
"loss": 3.196466979980469,
"step": 147500
},
{
"epoch": 0.3224370649707409,
"grad_norm": 0.1936854124069214,
"learning_rate": 0.003,
"loss": 3.1779425048828127,
"step": 147600
},
{
"epoch": 0.32265551826679156,
"grad_norm": 0.19209504127502441,
"learning_rate": 0.003,
"loss": 3.1815972900390626,
"step": 147700
},
{
"epoch": 0.3228739715628422,
"grad_norm": 0.1692599505186081,
"learning_rate": 0.003,
"loss": 3.185713195800781,
"step": 147800
},
{
"epoch": 0.32309242485889283,
"grad_norm": 0.19620276987552643,
"learning_rate": 0.003,
"loss": 3.1745257568359375,
"step": 147900
},
{
"epoch": 0.32331087815494347,
"grad_norm": 0.23885881900787354,
"learning_rate": 0.003,
"loss": 3.13314697265625,
"step": 148000
},
{
"epoch": 0.3235293314509941,
"grad_norm": 0.15670974552631378,
"learning_rate": 0.003,
"loss": 3.141082763671875,
"step": 148100
},
{
"epoch": 0.32374778474704474,
"grad_norm": 0.17646116018295288,
"learning_rate": 0.003,
"loss": 3.17531982421875,
"step": 148200
},
{
"epoch": 0.3239662380430954,
"grad_norm": 0.23557858169078827,
"learning_rate": 0.003,
"loss": 3.19098388671875,
"step": 148300
},
{
"epoch": 0.324184691339146,
"grad_norm": 0.14398302137851715,
"learning_rate": 0.003,
"loss": 3.1819873046875,
"step": 148400
},
{
"epoch": 0.32440314463519665,
"grad_norm": 0.18726244568824768,
"learning_rate": 0.003,
"loss": 3.1925845336914063,
"step": 148500
},
{
"epoch": 0.3246215979312473,
"grad_norm": 0.2285359650850296,
"learning_rate": 0.003,
"loss": 3.20674072265625,
"step": 148600
},
{
"epoch": 0.3248400512272979,
"grad_norm": 0.17976802587509155,
"learning_rate": 0.003,
"loss": 3.173955993652344,
"step": 148700
},
{
"epoch": 0.32505850452334856,
"grad_norm": 0.1975952386856079,
"learning_rate": 0.003,
"loss": 2.6845458984375,
"step": 148800
},
{
"epoch": 0.3252769578193992,
"grad_norm": 0.15447331964969635,
"learning_rate": 0.003,
"loss": 3.1710690307617186,
"step": 148900
},
{
"epoch": 0.32549541111544983,
"grad_norm": 0.18206952512264252,
"learning_rate": 0.003,
"loss": 3.158231506347656,
"step": 149000
},
{
"epoch": 0.32571386441150046,
"grad_norm": 0.17522966861724854,
"learning_rate": 0.003,
"loss": 3.1928945922851564,
"step": 149100
},
{
"epoch": 0.3259323177075511,
"grad_norm": 0.16397172212600708,
"learning_rate": 0.003,
"loss": 3.2015060424804687,
"step": 149200
},
{
"epoch": 0.32615077100360174,
"grad_norm": 0.21215063333511353,
"learning_rate": 0.003,
"loss": 3.1936944580078124,
"step": 149300
},
{
"epoch": 0.32636922429965237,
"grad_norm": 0.2076343297958374,
"learning_rate": 0.003,
"loss": 2.6863333129882814,
"step": 149400
},
{
"epoch": 0.326587677595703,
"grad_norm": 0.2564661502838135,
"learning_rate": 0.003,
"loss": 3.208557434082031,
"step": 149500
},
{
"epoch": 0.32680613089175364,
"grad_norm": 0.18941468000411987,
"learning_rate": 0.003,
"loss": 3.137552490234375,
"step": 149600
},
{
"epoch": 0.3270245841878043,
"grad_norm": 0.1444561928510666,
"learning_rate": 0.003,
"loss": 3.248238830566406,
"step": 149700
},
{
"epoch": 0.3272430374838549,
"grad_norm": 0.26616400480270386,
"learning_rate": 0.003,
"loss": 3.21462890625,
"step": 149800
},
{
"epoch": 0.32746149077990555,
"grad_norm": 0.20099005103111267,
"learning_rate": 0.003,
"loss": 3.169850158691406,
"step": 149900
},
{
"epoch": 0.3276799440759562,
"grad_norm": 0.16536197066307068,
"learning_rate": 0.003,
"loss": 3.1830929565429686,
"step": 150000
},
{
"epoch": 0.3278983973720068,
"grad_norm": 0.18397554755210876,
"learning_rate": 0.003,
"loss": 3.2401638793945313,
"step": 150100
},
{
"epoch": 0.3281168506680575,
"grad_norm": 0.16635970771312714,
"learning_rate": 0.003,
"loss": 3.1912631225585937,
"step": 150200
},
{
"epoch": 0.32833530396410815,
"grad_norm": 0.1743297576904297,
"learning_rate": 0.003,
"loss": 3.2167935180664062,
"step": 150300
},
{
"epoch": 0.3285537572601588,
"grad_norm": 0.2537592947483063,
"learning_rate": 0.003,
"loss": 3.2288177490234373,
"step": 150400
},
{
"epoch": 0.3287722105562094,
"grad_norm": 0.1927787810564041,
"learning_rate": 0.003,
"loss": 3.170396423339844,
"step": 150500
},
{
"epoch": 0.32899066385226006,
"grad_norm": 0.23473043739795685,
"learning_rate": 0.003,
"loss": 3.1796908569335938,
"step": 150600
},
{
"epoch": 0.3292091171483107,
"grad_norm": 0.2488144338130951,
"learning_rate": 0.003,
"loss": 3.16794189453125,
"step": 150700
},
{
"epoch": 0.32942757044436133,
"grad_norm": 0.21751940250396729,
"learning_rate": 0.003,
"loss": 3.185949401855469,
"step": 150800
},
{
"epoch": 0.32964602374041196,
"grad_norm": 0.22387519478797913,
"learning_rate": 0.003,
"loss": 3.202431945800781,
"step": 150900
},
{
"epoch": 0.3298644770364626,
"grad_norm": 0.1522534191608429,
"learning_rate": 0.003,
"loss": 3.211038818359375,
"step": 151000
},
{
"epoch": 0.33008293033251324,
"grad_norm": 0.17218022048473358,
"learning_rate": 0.003,
"loss": 3.1591448974609375,
"step": 151100
},
{
"epoch": 0.33030138362856387,
"grad_norm": 0.17332923412322998,
"learning_rate": 0.003,
"loss": 3.147066955566406,
"step": 151200
},
{
"epoch": 0.3305198369246145,
"grad_norm": 0.16366907954216003,
"learning_rate": 0.003,
"loss": 3.228832092285156,
"step": 151300
},
{
"epoch": 0.33073829022066514,
"grad_norm": 0.18040701746940613,
"learning_rate": 0.003,
"loss": 3.154132080078125,
"step": 151400
},
{
"epoch": 0.3309567435167158,
"grad_norm": 0.17774927616119385,
"learning_rate": 0.003,
"loss": 3.192138671875,
"step": 151500
},
{
"epoch": 0.3311751968127664,
"grad_norm": 0.20536623895168304,
"learning_rate": 0.003,
"loss": 3.18081787109375,
"step": 151600
},
{
"epoch": 0.33139365010881705,
"grad_norm": 0.20949424803256989,
"learning_rate": 0.003,
"loss": 3.1637872314453124,
"step": 151700
},
{
"epoch": 0.3316121034048677,
"grad_norm": 0.2223246991634369,
"learning_rate": 0.003,
"loss": 3.1776608276367186,
"step": 151800
},
{
"epoch": 0.3318305567009183,
"grad_norm": 0.17206287384033203,
"learning_rate": 0.003,
"loss": 3.15651123046875,
"step": 151900
},
{
"epoch": 0.33204900999696896,
"grad_norm": 0.21758702397346497,
"learning_rate": 0.003,
"loss": 3.1671078491210936,
"step": 152000
},
{
"epoch": 0.3322674632930196,
"grad_norm": 0.18254788219928741,
"learning_rate": 0.003,
"loss": 3.1718585205078127,
"step": 152100
},
{
"epoch": 0.33248591658907023,
"grad_norm": 0.18393871188163757,
"learning_rate": 0.003,
"loss": 3.1879238891601562,
"step": 152200
},
{
"epoch": 0.33270436988512087,
"grad_norm": 0.23279863595962524,
"learning_rate": 0.003,
"loss": 3.192552185058594,
"step": 152300
},
{
"epoch": 0.3329228231811715,
"grad_norm": 0.18877536058425903,
"learning_rate": 0.003,
"loss": 3.177838134765625,
"step": 152400
},
{
"epoch": 0.33314127647722214,
"grad_norm": 0.1876782476902008,
"learning_rate": 0.003,
"loss": 3.132906494140625,
"step": 152500
},
{
"epoch": 0.3333597297732728,
"grad_norm": 0.16887180507183075,
"learning_rate": 0.003,
"loss": 3.2051397705078126,
"step": 152600
},
{
"epoch": 0.3335781830693234,
"grad_norm": 0.1972004771232605,
"learning_rate": 0.003,
"loss": 3.1865728759765624,
"step": 152700
},
{
"epoch": 0.33379663636537404,
"grad_norm": 0.2580986022949219,
"learning_rate": 0.003,
"loss": 3.2121429443359375,
"step": 152800
},
{
"epoch": 0.3340150896614247,
"grad_norm": 0.1952686458826065,
"learning_rate": 0.003,
"loss": 3.1715158081054686,
"step": 152900
},
{
"epoch": 0.3342335429574753,
"grad_norm": 0.22782842814922333,
"learning_rate": 0.003,
"loss": 3.1909320068359377,
"step": 153000
},
{
"epoch": 0.33445199625352595,
"grad_norm": 0.2532144784927368,
"learning_rate": 0.003,
"loss": 3.2305587768554687,
"step": 153100
},
{
"epoch": 0.3346704495495766,
"grad_norm": 0.22145599126815796,
"learning_rate": 0.003,
"loss": 3.2270474243164062,
"step": 153200
},
{
"epoch": 0.3348889028456272,
"grad_norm": 0.21716009080410004,
"learning_rate": 0.003,
"loss": 3.1774713134765626,
"step": 153300
},
{
"epoch": 0.33510735614167786,
"grad_norm": 0.31625574827194214,
"learning_rate": 0.003,
"loss": 2.6925650024414063,
"step": 153400
},
{
"epoch": 0.33532580943772855,
"grad_norm": 0.18389812111854553,
"learning_rate": 0.003,
"loss": 3.2025628662109376,
"step": 153500
},
{
"epoch": 0.3355442627337792,
"grad_norm": 0.18168075382709503,
"learning_rate": 0.003,
"loss": 3.1919210815429686,
"step": 153600
},
{
"epoch": 0.3357627160298298,
"grad_norm": 0.19095513224601746,
"learning_rate": 0.003,
"loss": 3.170085754394531,
"step": 153700
},
{
"epoch": 0.33598116932588046,
"grad_norm": 0.24979139864444733,
"learning_rate": 0.003,
"loss": 3.196229248046875,
"step": 153800
},
{
"epoch": 0.3361996226219311,
"grad_norm": 0.24247103929519653,
"learning_rate": 0.003,
"loss": 3.155550231933594,
"step": 153900
},
{
"epoch": 0.33641807591798173,
"grad_norm": 0.2211730033159256,
"learning_rate": 0.003,
"loss": 3.2138156127929687,
"step": 154000
},
{
"epoch": 0.33663652921403237,
"grad_norm": 0.1641773283481598,
"learning_rate": 0.003,
"loss": 3.1940878295898436,
"step": 154100
},
{
"epoch": 0.336854982510083,
"grad_norm": 0.1894778609275818,
"learning_rate": 0.003,
"loss": 3.2004910278320313,
"step": 154200
},
{
"epoch": 0.33707343580613364,
"grad_norm": 0.1566590666770935,
"learning_rate": 0.003,
"loss": 3.1311114501953123,
"step": 154300
},
{
"epoch": 0.3372918891021843,
"grad_norm": 0.19558022916316986,
"learning_rate": 0.003,
"loss": 3.164029235839844,
"step": 154400
},
{
"epoch": 0.3375103423982349,
"grad_norm": 0.21856990456581116,
"learning_rate": 0.003,
"loss": 3.1604824829101563,
"step": 154500
},
{
"epoch": 0.33772879569428554,
"grad_norm": 0.19755780696868896,
"learning_rate": 0.003,
"loss": 3.158997802734375,
"step": 154600
},
{
"epoch": 0.3379472489903362,
"grad_norm": 0.15918242931365967,
"learning_rate": 0.003,
"loss": 3.230296325683594,
"step": 154700
},
{
"epoch": 0.3381657022863868,
"grad_norm": 0.24871376156806946,
"learning_rate": 0.003,
"loss": 2.9430838012695313,
"step": 154800
},
{
"epoch": 0.33838415558243745,
"grad_norm": 0.24803753197193146,
"learning_rate": 0.003,
"loss": 3.2192874145507813,
"step": 154900
},
{
"epoch": 0.3386026088784881,
"grad_norm": 0.17800495028495789,
"learning_rate": 0.003,
"loss": 3.204012756347656,
"step": 155000
},
{
"epoch": 0.3388210621745387,
"grad_norm": 0.25822532176971436,
"learning_rate": 0.003,
"loss": 3.18099365234375,
"step": 155100
},
{
"epoch": 0.33903951547058936,
"grad_norm": 0.16437004506587982,
"learning_rate": 0.003,
"loss": 3.183768310546875,
"step": 155200
},
{
"epoch": 0.33925796876664,
"grad_norm": 0.1815718412399292,
"learning_rate": 0.003,
"loss": 3.076330261230469,
"step": 155300
},
{
"epoch": 0.33947642206269063,
"grad_norm": 0.1795693188905716,
"learning_rate": 0.003,
"loss": 3.1956195068359374,
"step": 155400
},
{
"epoch": 0.33969487535874127,
"grad_norm": 0.9455909729003906,
"learning_rate": 0.003,
"loss": 3.154691467285156,
"step": 155500
},
{
"epoch": 0.3399133286547919,
"grad_norm": 0.188655823469162,
"learning_rate": 0.003,
"loss": 3.1824188232421875,
"step": 155600
},
{
"epoch": 0.34013178195084254,
"grad_norm": 0.19210544228553772,
"learning_rate": 0.003,
"loss": 3.18107177734375,
"step": 155700
},
{
"epoch": 0.3403502352468932,
"grad_norm": 0.16656406223773956,
"learning_rate": 0.003,
"loss": 3.1568673706054686,
"step": 155800
},
{
"epoch": 0.3405686885429438,
"grad_norm": 0.22872625291347504,
"learning_rate": 0.003,
"loss": 3.171075744628906,
"step": 155900
},
{
"epoch": 0.34078714183899445,
"grad_norm": 0.16131971776485443,
"learning_rate": 0.003,
"loss": 3.167562255859375,
"step": 156000
},
{
"epoch": 0.3410055951350451,
"grad_norm": 0.15685568749904633,
"learning_rate": 0.003,
"loss": 3.1865103149414065,
"step": 156100
},
{
"epoch": 0.3412240484310957,
"grad_norm": 0.22560730576515198,
"learning_rate": 0.003,
"loss": 3.208013916015625,
"step": 156200
},
{
"epoch": 0.34144250172714635,
"grad_norm": 0.1931314617395401,
"learning_rate": 0.003,
"loss": 3.1950698852539063,
"step": 156300
},
{
"epoch": 0.341660955023197,
"grad_norm": 0.1877717822790146,
"learning_rate": 0.003,
"loss": 3.1804913330078124,
"step": 156400
},
{
"epoch": 0.3418794083192476,
"grad_norm": 0.2338135540485382,
"learning_rate": 0.003,
"loss": 3.1736178588867188,
"step": 156500
},
{
"epoch": 0.34209786161529826,
"grad_norm": 0.2008652240037918,
"learning_rate": 0.003,
"loss": 3.1773001098632814,
"step": 156600
},
{
"epoch": 0.3423163149113489,
"grad_norm": 0.17682594060897827,
"learning_rate": 0.003,
"loss": 3.19855712890625,
"step": 156700
},
{
"epoch": 0.3425347682073996,
"grad_norm": 0.1595831662416458,
"learning_rate": 0.003,
"loss": 3.1685079956054687,
"step": 156800
},
{
"epoch": 0.3427532215034502,
"grad_norm": 0.14333690702915192,
"learning_rate": 0.003,
"loss": 3.187269592285156,
"step": 156900
},
{
"epoch": 0.34297167479950086,
"grad_norm": 0.2140931487083435,
"learning_rate": 0.003,
"loss": 3.1413455200195313,
"step": 157000
},
{
"epoch": 0.3431901280955515,
"grad_norm": 0.22988301515579224,
"learning_rate": 0.003,
"loss": 3.222878723144531,
"step": 157100
},
{
"epoch": 0.34340858139160213,
"grad_norm": 0.1641242504119873,
"learning_rate": 0.003,
"loss": 3.16712890625,
"step": 157200
},
{
"epoch": 0.34362703468765277,
"grad_norm": 0.1825564205646515,
"learning_rate": 0.003,
"loss": 3.16823486328125,
"step": 157300
},
{
"epoch": 0.3438454879837034,
"grad_norm": 0.18540945649147034,
"learning_rate": 0.003,
"loss": 3.1882949829101563,
"step": 157400
},
{
"epoch": 0.34406394127975404,
"grad_norm": 0.2009730190038681,
"learning_rate": 0.003,
"loss": 3.2248907470703125,
"step": 157500
},
{
"epoch": 0.3442823945758047,
"grad_norm": 0.17075271904468536,
"learning_rate": 0.003,
"loss": 3.0380267333984374,
"step": 157600
},
{
"epoch": 0.3445008478718553,
"grad_norm": 0.1846093088388443,
"learning_rate": 0.003,
"loss": 3.610770263671875,
"step": 157700
},
{
"epoch": 0.34471930116790594,
"grad_norm": 0.16304230690002441,
"learning_rate": 0.003,
"loss": 3.1604934692382813,
"step": 157800
},
{
"epoch": 0.3449377544639566,
"grad_norm": 0.19401445984840393,
"learning_rate": 0.003,
"loss": 3.1814840698242186,
"step": 157900
},
{
"epoch": 0.3451562077600072,
"grad_norm": 0.15682169795036316,
"learning_rate": 0.003,
"loss": 3.1792025756835938,
"step": 158000
},
{
"epoch": 0.34537466105605785,
"grad_norm": 0.21715964376926422,
"learning_rate": 0.003,
"loss": 3.2103289794921874,
"step": 158100
},
{
"epoch": 0.3455931143521085,
"grad_norm": 0.167643204331398,
"learning_rate": 0.003,
"loss": 3.1870416259765624,
"step": 158200
},
{
"epoch": 0.3458115676481591,
"grad_norm": 0.21195171773433685,
"learning_rate": 0.003,
"loss": 3.2112298583984376,
"step": 158300
},
{
"epoch": 0.34603002094420976,
"grad_norm": 0.20956410467624664,
"learning_rate": 0.003,
"loss": 3.1495452880859376,
"step": 158400
},
{
"epoch": 0.3462484742402604,
"grad_norm": 0.17523027956485748,
"learning_rate": 0.003,
"loss": 3.1977639770507813,
"step": 158500
},
{
"epoch": 0.34646692753631103,
"grad_norm": 0.1980908364057541,
"learning_rate": 0.003,
"loss": 3.1888766479492188,
"step": 158600
},
{
"epoch": 0.34668538083236167,
"grad_norm": 0.14576759934425354,
"learning_rate": 0.003,
"loss": 3.16719482421875,
"step": 158700
},
{
"epoch": 0.3469038341284123,
"grad_norm": 0.16576829552650452,
"learning_rate": 0.003,
"loss": 3.214666442871094,
"step": 158800
},
{
"epoch": 0.34712228742446294,
"grad_norm": 0.16519513726234436,
"learning_rate": 0.003,
"loss": 3.1716073608398436,
"step": 158900
},
{
"epoch": 0.3473407407205136,
"grad_norm": 0.24041765928268433,
"learning_rate": 0.003,
"loss": 3.1977297973632814,
"step": 159000
},
{
"epoch": 0.3475591940165642,
"grad_norm": 0.20679371058940887,
"learning_rate": 0.003,
"loss": 3.1427734375,
"step": 159100
},
{
"epoch": 0.34777764731261485,
"grad_norm": 0.13883677124977112,
"learning_rate": 0.003,
"loss": 3.21267333984375,
"step": 159200
},
{
"epoch": 0.3479961006086655,
"grad_norm": 0.26354363560676575,
"learning_rate": 0.003,
"loss": 3.2137460327148437,
"step": 159300
},
{
"epoch": 0.3482145539047161,
"grad_norm": 0.20705294609069824,
"learning_rate": 0.003,
"loss": 3.1597760009765623,
"step": 159400
},
{
"epoch": 0.34843300720076675,
"grad_norm": 0.22643183171749115,
"learning_rate": 0.003,
"loss": 3.167184143066406,
"step": 159500
},
{
"epoch": 0.3486514604968174,
"grad_norm": 0.18660259246826172,
"learning_rate": 0.003,
"loss": 2.986766052246094,
"step": 159600
},
{
"epoch": 0.348869913792868,
"grad_norm": 0.15333609282970428,
"learning_rate": 0.003,
"loss": 3.2087274169921876,
"step": 159700
},
{
"epoch": 0.34908836708891866,
"grad_norm": 0.22221969068050385,
"learning_rate": 0.003,
"loss": 3.1692669677734373,
"step": 159800
},
{
"epoch": 0.3493068203849693,
"grad_norm": 0.22474493086338043,
"learning_rate": 0.003,
"loss": 3.1715362548828123,
"step": 159900
},
{
"epoch": 0.34952527368101993,
"grad_norm": 0.198649063706398,
"learning_rate": 0.003,
"loss": 3.157160339355469,
"step": 160000
},
{
"epoch": 0.3497437269770706,
"grad_norm": 0.15874658524990082,
"learning_rate": 0.003,
"loss": 3.153104553222656,
"step": 160100
},
{
"epoch": 0.34996218027312126,
"grad_norm": 0.16205880045890808,
"learning_rate": 0.003,
"loss": 3.1441021728515626,
"step": 160200
},
{
"epoch": 0.3501806335691719,
"grad_norm": 0.15748894214630127,
"learning_rate": 0.003,
"loss": 3.1915768432617186,
"step": 160300
},
{
"epoch": 0.35039908686522253,
"grad_norm": 0.16804157197475433,
"learning_rate": 0.003,
"loss": 2.852764892578125,
"step": 160400
},
{
"epoch": 0.35061754016127317,
"grad_norm": 0.1834060698747635,
"learning_rate": 0.003,
"loss": 3.20140380859375,
"step": 160500
},
{
"epoch": 0.3508359934573238,
"grad_norm": 0.17118389904499054,
"learning_rate": 0.003,
"loss": 3.1939096069335937,
"step": 160600
},
{
"epoch": 0.35105444675337444,
"grad_norm": 0.1672763228416443,
"learning_rate": 0.003,
"loss": 3.1681689453125,
"step": 160700
},
{
"epoch": 0.3512729000494251,
"grad_norm": 0.19242221117019653,
"learning_rate": 0.003,
"loss": 3.1664584350585936,
"step": 160800
},
{
"epoch": 0.3514913533454757,
"grad_norm": 0.15714991092681885,
"learning_rate": 0.003,
"loss": 3.109347229003906,
"step": 160900
},
{
"epoch": 0.35170980664152635,
"grad_norm": 0.15889999270439148,
"learning_rate": 0.003,
"loss": 3.2017071533203123,
"step": 161000
},
{
"epoch": 0.351928259937577,
"grad_norm": 0.1645916849374771,
"learning_rate": 0.003,
"loss": 3.167985534667969,
"step": 161100
},
{
"epoch": 0.3521467132336276,
"grad_norm": 0.23582269251346588,
"learning_rate": 0.003,
"loss": 3.177903137207031,
"step": 161200
},
{
"epoch": 0.35236516652967825,
"grad_norm": 0.19157132506370544,
"learning_rate": 0.003,
"loss": 3.191995849609375,
"step": 161300
},
{
"epoch": 0.3525836198257289,
"grad_norm": 0.20941288769245148,
"learning_rate": 0.003,
"loss": 3.153175964355469,
"step": 161400
},
{
"epoch": 0.3528020731217795,
"grad_norm": 0.1850237399339676,
"learning_rate": 0.003,
"loss": 3.170132751464844,
"step": 161500
},
{
"epoch": 0.35302052641783016,
"grad_norm": 0.16023872792720795,
"learning_rate": 0.003,
"loss": 3.1849407958984375,
"step": 161600
},
{
"epoch": 0.3532389797138808,
"grad_norm": 0.16243669390678406,
"learning_rate": 0.003,
"loss": 3.1426345825195314,
"step": 161700
},
{
"epoch": 0.35345743300993143,
"grad_norm": 0.18923358619213104,
"learning_rate": 0.003,
"loss": 3.1663934326171876,
"step": 161800
},
{
"epoch": 0.35367588630598207,
"grad_norm": 0.1994985044002533,
"learning_rate": 0.003,
"loss": 3.2043157958984376,
"step": 161900
},
{
"epoch": 0.3538943396020327,
"grad_norm": 0.19438615441322327,
"learning_rate": 0.003,
"loss": 3.2031292724609375,
"step": 162000
},
{
"epoch": 0.35411279289808334,
"grad_norm": 0.1888693869113922,
"learning_rate": 0.003,
"loss": 3.174179992675781,
"step": 162100
},
{
"epoch": 0.354331246194134,
"grad_norm": 0.20153461396694183,
"learning_rate": 0.003,
"loss": 3.2011062622070314,
"step": 162200
},
{
"epoch": 0.3545496994901846,
"grad_norm": 0.25670602917671204,
"learning_rate": 0.003,
"loss": 3.2489642333984374,
"step": 162300
},
{
"epoch": 0.35476815278623525,
"grad_norm": 0.3643508851528168,
"learning_rate": 0.003,
"loss": 3.1777423095703123,
"step": 162400
},
{
"epoch": 0.3549866060822859,
"grad_norm": 0.14907072484493256,
"learning_rate": 0.003,
"loss": 3.182525634765625,
"step": 162500
},
{
"epoch": 0.3552050593783365,
"grad_norm": 0.18209069967269897,
"learning_rate": 0.003,
"loss": 3.164674072265625,
"step": 162600
},
{
"epoch": 0.35542351267438715,
"grad_norm": 0.2017740160226822,
"learning_rate": 0.003,
"loss": 3.194468688964844,
"step": 162700
},
{
"epoch": 0.3556419659704378,
"grad_norm": 0.17735159397125244,
"learning_rate": 0.003,
"loss": 3.179418029785156,
"step": 162800
},
{
"epoch": 0.3558604192664884,
"grad_norm": 0.25465652346611023,
"learning_rate": 0.003,
"loss": 3.181459655761719,
"step": 162900
},
{
"epoch": 0.35607887256253906,
"grad_norm": 0.15170180797576904,
"learning_rate": 0.003,
"loss": 3.1619601440429688,
"step": 163000
},
{
"epoch": 0.3562973258585897,
"grad_norm": 0.18672648072242737,
"learning_rate": 0.003,
"loss": 3.165998840332031,
"step": 163100
},
{
"epoch": 0.35651577915464033,
"grad_norm": 0.21018573641777039,
"learning_rate": 0.003,
"loss": 3.1902142333984376,
"step": 163200
},
{
"epoch": 0.35673423245069097,
"grad_norm": 0.5577970743179321,
"learning_rate": 0.003,
"loss": 3.1940487670898436,
"step": 163300
},
{
"epoch": 0.35695268574674166,
"grad_norm": 0.22371803224086761,
"learning_rate": 0.003,
"loss": 3.1257666015625,
"step": 163400
},
{
"epoch": 0.3571711390427923,
"grad_norm": 0.17022310197353363,
"learning_rate": 0.003,
"loss": 3.1521151733398436,
"step": 163500
},
{
"epoch": 0.35738959233884293,
"grad_norm": 0.16075775027275085,
"learning_rate": 0.003,
"loss": 3.1516033935546877,
"step": 163600
},
{
"epoch": 0.35760804563489357,
"grad_norm": 0.16093774139881134,
"learning_rate": 0.003,
"loss": 3.1644271850585937,
"step": 163700
},
{
"epoch": 0.3578264989309442,
"grad_norm": 0.20827049016952515,
"learning_rate": 0.003,
"loss": 3.182415771484375,
"step": 163800
},
{
"epoch": 0.35804495222699484,
"grad_norm": 0.17943401634693146,
"learning_rate": 0.003,
"loss": 3.1575823974609376,
"step": 163900
},
{
"epoch": 0.3582634055230455,
"grad_norm": 0.19859860837459564,
"learning_rate": 0.003,
"loss": 3.145097351074219,
"step": 164000
},
{
"epoch": 0.3584818588190961,
"grad_norm": 0.20336900651454926,
"learning_rate": 0.003,
"loss": 3.185714111328125,
"step": 164100
},
{
"epoch": 0.35870031211514675,
"grad_norm": 0.18309247493743896,
"learning_rate": 0.003,
"loss": 3.1826467895507813,
"step": 164200
},
{
"epoch": 0.3589187654111974,
"grad_norm": 0.17939050495624542,
"learning_rate": 0.003,
"loss": 3.1910614013671874,
"step": 164300
},
{
"epoch": 0.359137218707248,
"grad_norm": 0.20175869762897491,
"learning_rate": 0.003,
"loss": 3.1837725830078125,
"step": 164400
},
{
"epoch": 0.35935567200329865,
"grad_norm": 0.21120579540729523,
"learning_rate": 0.003,
"loss": 3.1697833251953127,
"step": 164500
},
{
"epoch": 0.3595741252993493,
"grad_norm": 0.1523100584745407,
"learning_rate": 0.003,
"loss": 3.161806945800781,
"step": 164600
},
{
"epoch": 0.3597925785953999,
"grad_norm": 0.18864744901657104,
"learning_rate": 0.003,
"loss": 3.1357794189453125,
"step": 164700
},
{
"epoch": 0.36001103189145056,
"grad_norm": 0.1667052060365677,
"learning_rate": 0.003,
"loss": 3.1624359130859374,
"step": 164800
},
{
"epoch": 0.3602294851875012,
"grad_norm": 0.5051073431968689,
"learning_rate": 0.003,
"loss": 3.1776788330078123,
"step": 164900
},
{
"epoch": 0.36044793848355183,
"grad_norm": 0.1835436075925827,
"learning_rate": 0.003,
"loss": 3.2048748779296874,
"step": 165000
},
{
"epoch": 0.36066639177960247,
"grad_norm": 0.2018285095691681,
"learning_rate": 0.003,
"loss": 3.1898895263671876,
"step": 165100
},
{
"epoch": 0.3608848450756531,
"grad_norm": 0.2360716611146927,
"learning_rate": 0.003,
"loss": 3.1715792846679687,
"step": 165200
},
{
"epoch": 0.36110329837170374,
"grad_norm": 0.19357648491859436,
"learning_rate": 0.003,
"loss": 3.1751763916015623,
"step": 165300
},
{
"epoch": 0.3613217516677544,
"grad_norm": 0.22794722020626068,
"learning_rate": 0.003,
"loss": 3.1800875854492188,
"step": 165400
},
{
"epoch": 0.361540204963805,
"grad_norm": 0.18184566497802734,
"learning_rate": 0.003,
"loss": 3.2099285888671876,
"step": 165500
},
{
"epoch": 0.36175865825985565,
"grad_norm": 0.1959710717201233,
"learning_rate": 0.003,
"loss": 3.0024569702148436,
"step": 165600
},
{
"epoch": 0.3619771115559063,
"grad_norm": 0.2441798746585846,
"learning_rate": 0.003,
"loss": 3.160757751464844,
"step": 165700
},
{
"epoch": 0.3621955648519569,
"grad_norm": 0.2302255481481552,
"learning_rate": 0.003,
"loss": 3.1961187744140624,
"step": 165800
},
{
"epoch": 0.36241401814800756,
"grad_norm": 0.22491635382175446,
"learning_rate": 0.003,
"loss": 3.1815185546875,
"step": 165900
},
{
"epoch": 0.3626324714440582,
"grad_norm": 0.19608400762081146,
"learning_rate": 0.003,
"loss": 3.2075823974609374,
"step": 166000
},
{
"epoch": 0.3628509247401088,
"grad_norm": 0.18288902938365936,
"learning_rate": 0.003,
"loss": 3.1492605590820313,
"step": 166100
},
{
"epoch": 0.36306937803615946,
"grad_norm": 0.20123060047626495,
"learning_rate": 0.003,
"loss": 3.1266104125976564,
"step": 166200
},
{
"epoch": 0.3632878313322101,
"grad_norm": 0.16410323977470398,
"learning_rate": 0.003,
"loss": 3.211439208984375,
"step": 166300
},
{
"epoch": 0.36350628462826073,
"grad_norm": 0.27165287733078003,
"learning_rate": 0.003,
"loss": 3.137961730957031,
"step": 166400
},
{
"epoch": 0.36372473792431137,
"grad_norm": 0.17709319293498993,
"learning_rate": 0.003,
"loss": 3.179796142578125,
"step": 166500
},
{
"epoch": 0.363943191220362,
"grad_norm": 0.1782919019460678,
"learning_rate": 0.003,
"loss": 3.16959716796875,
"step": 166600
},
{
"epoch": 0.3641616445164127,
"grad_norm": 0.35208815336227417,
"learning_rate": 0.003,
"loss": 3.157923583984375,
"step": 166700
},
{
"epoch": 0.36438009781246333,
"grad_norm": 0.2911792993545532,
"learning_rate": 0.003,
"loss": 3.1828079223632812,
"step": 166800
},
{
"epoch": 0.36459855110851397,
"grad_norm": 0.1510099470615387,
"learning_rate": 0.003,
"loss": 3.1355581665039063,
"step": 166900
},
{
"epoch": 0.3648170044045646,
"grad_norm": 0.15822523832321167,
"learning_rate": 0.003,
"loss": 3.156095275878906,
"step": 167000
},
{
"epoch": 0.36503545770061524,
"grad_norm": 0.21178464591503143,
"learning_rate": 0.003,
"loss": 3.1742117309570315,
"step": 167100
},
{
"epoch": 0.3652539109966659,
"grad_norm": 0.1757752001285553,
"learning_rate": 0.003,
"loss": 3.174485778808594,
"step": 167200
},
{
"epoch": 0.3654723642927165,
"grad_norm": 0.18174314498901367,
"learning_rate": 0.003,
"loss": 3.14523193359375,
"step": 167300
},
{
"epoch": 0.36569081758876715,
"grad_norm": 0.518521249294281,
"learning_rate": 0.003,
"loss": 3.1884396362304686,
"step": 167400
},
{
"epoch": 0.3659092708848178,
"grad_norm": 0.16352346539497375,
"learning_rate": 0.003,
"loss": 3.2025137329101563,
"step": 167500
},
{
"epoch": 0.3661277241808684,
"grad_norm": 0.18585847318172455,
"learning_rate": 0.003,
"loss": 3.1688381958007814,
"step": 167600
},
{
"epoch": 0.36634617747691905,
"grad_norm": 0.1814800202846527,
"learning_rate": 0.003,
"loss": 3.1297698974609376,
"step": 167700
},
{
"epoch": 0.3665646307729697,
"grad_norm": 0.21663619577884674,
"learning_rate": 0.003,
"loss": 3.1845257568359373,
"step": 167800
},
{
"epoch": 0.3667830840690203,
"grad_norm": 0.19967739284038544,
"learning_rate": 0.003,
"loss": 3.168363037109375,
"step": 167900
},
{
"epoch": 0.36700153736507096,
"grad_norm": 0.21710465848445892,
"learning_rate": 0.003,
"loss": 3.141382141113281,
"step": 168000
}
],
"logging_steps": 100,
"max_steps": 457764,
"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": false
},
"attributes": {}
}
},
"total_flos": 3.29248703250432e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}