Files
Oolel-Corrector/trainer_state.json

1155 lines
30 KiB
JSON
Raw Normal View History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 13953,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0002150422020321488,
"grad_norm": 12.0,
"learning_rate": 1.4326647564469915e-07,
"loss": 1.6164790391921997,
"step": 1,
"token_acc": 0.6621004566210046
},
{
"epoch": 0.02150422020321488,
"grad_norm": 6.21875,
"learning_rate": 1.4326647564469913e-05,
"loss": 1.188794762197167,
"step": 100,
"token_acc": 0.7460876453083856
},
{
"epoch": 0.04300844040642976,
"grad_norm": 4.46875,
"learning_rate": 2.8653295128939826e-05,
"loss": 0.7355960845947266,
"step": 200,
"token_acc": 0.8187009896706506
},
{
"epoch": 0.06451266060964464,
"grad_norm": 2.78125,
"learning_rate": 4.297994269340974e-05,
"loss": 0.5316849899291992,
"step": 300,
"token_acc": 0.8615559780448769
},
{
"epoch": 0.08601688081285952,
"grad_norm": 3.75,
"learning_rate": 5.730659025787965e-05,
"loss": 0.46149860382080077,
"step": 400,
"token_acc": 0.8763085964931234
},
{
"epoch": 0.1075211010160744,
"grad_norm": 3.046875,
"learning_rate": 7.163323782234958e-05,
"loss": 0.43257137298583986,
"step": 500,
"token_acc": 0.8855033733461842
},
{
"epoch": 0.12902532121928928,
"grad_norm": 2.828125,
"learning_rate": 8.595988538681948e-05,
"loss": 0.4290001678466797,
"step": 600,
"token_acc": 0.8841237425308222
},
{
"epoch": 0.15052954142250416,
"grad_norm": 2.84375,
"learning_rate": 9.999999438253633e-05,
"loss": 0.4132740020751953,
"step": 700,
"token_acc": 0.8884292300890545
},
{
"epoch": 0.17203376162571904,
"grad_norm": 2.765625,
"learning_rate": 9.998538968828853e-05,
"loss": 0.4122954940795898,
"step": 800,
"token_acc": 0.8891312971043237
},
{
"epoch": 0.19353798182893392,
"grad_norm": 2.4375,
"learning_rate": 9.994270719682668e-05,
"loss": 0.41482566833496093,
"step": 900,
"token_acc": 0.8889188460501483
},
{
"epoch": 0.2150422020321488,
"grad_norm": 3.203125,
"learning_rate": 9.987197088376343e-05,
"loss": 0.3782763671875,
"step": 1000,
"token_acc": 0.8961677752757686
},
{
"epoch": 0.23654642223536368,
"grad_norm": 2.421875,
"learning_rate": 9.977322048310636e-05,
"loss": 0.38466773986816405,
"step": 1100,
"token_acc": 0.8960082933674004
},
{
"epoch": 0.25805064243857856,
"grad_norm": 1.8828125,
"learning_rate": 9.964651146493868e-05,
"loss": 0.37627601623535156,
"step": 1200,
"token_acc": 0.8976215770873534
},
{
"epoch": 0.27955486264179347,
"grad_norm": 2.0,
"learning_rate": 9.94919150042605e-05,
"loss": 0.37790138244628907,
"step": 1300,
"token_acc": 0.8967620047231698
},
{
"epoch": 0.3010590828450083,
"grad_norm": 1.7109375,
"learning_rate": 9.930951794100846e-05,
"loss": 0.3538982009887695,
"step": 1400,
"token_acc": 0.9031133435062733
},
{
"epoch": 0.3225633030482232,
"grad_norm": 1.609375,
"learning_rate": 9.909942273127595e-05,
"loss": 0.35670711517333986,
"step": 1500,
"token_acc": 0.9019035105816946
},
{
"epoch": 0.3440675232514381,
"grad_norm": 2.296875,
"learning_rate": 9.886174738976148e-05,
"loss": 0.34810089111328124,
"step": 1600,
"token_acc": 0.9055337820567436
},
{
"epoch": 0.365571743454653,
"grad_norm": 2.015625,
"learning_rate": 9.85966254234775e-05,
"loss": 0.3589338302612305,
"step": 1700,
"token_acc": 0.9023834266367202
},
{
"epoch": 0.38707596365786784,
"grad_norm": 1.875,
"learning_rate": 9.830420575675672e-05,
"loss": 0.3577522659301758,
"step": 1800,
"token_acc": 0.9015066534697019
},
{
"epoch": 0.40858018386108275,
"grad_norm": 1.8359375,
"learning_rate": 9.798465264759846e-05,
"loss": 0.34096073150634765,
"step": 1900,
"token_acc": 0.905558431474052
},
{
"epoch": 0.4300844040642976,
"grad_norm": 2.0625,
"learning_rate": 9.763814559540149e-05,
"loss": 0.3458216094970703,
"step": 2000,
"token_acc": 0.9044097244455828
},
{
"epoch": 0.4515886242675125,
"grad_norm": 2.53125,
"learning_rate": 9.726487924013573e-05,
"loss": 0.34719383239746093,
"step": 2100,
"token_acc": 0.9023519529609407
},
{
"epoch": 0.47309284447072736,
"grad_norm": 2.078125,
"learning_rate": 9.686506325300902e-05,
"loss": 0.33799285888671876,
"step": 2200,
"token_acc": 0.9068587312532322
},
{
"epoch": 0.49459706467394227,
"grad_norm": 1.890625,
"learning_rate": 9.643892221869059e-05,
"loss": 0.3353384780883789,
"step": 2300,
"token_acc": 0.9076656738244497
},
{
"epoch": 0.5161012848771571,
"grad_norm": 1.5546875,
"learning_rate": 9.598669550915733e-05,
"loss": 0.3250057601928711,
"step": 2400,
"token_acc": 0.9099034988409078
},
{
"epoch": 0.537605505080372,
"grad_norm": 1.859375,
"learning_rate": 9.55086371492338e-05,
"loss": 0.32469905853271486,
"step": 2500,
"token_acc": 0.9092186712153885
},
{
"epoch": 0.5591097252835869,
"grad_norm": 1.2578125,
"learning_rate": 9.500501567390132e-05,
"loss": 0.33898296356201174,
"step": 2600,
"token_acc": 0.9060987206085753
},
{
"epoch": 0.5806139454868018,
"grad_norm": 1.7265625,
"learning_rate": 9.447611397745651e-05,
"loss": 0.32965770721435544,
"step": 2700,
"token_acc": 0.9072501632919661
},
{
"epoch": 0.6021181656900166,
"grad_norm": 2.25,
"learning_rate": 9.392222915460385e-05,
"loss": 0.3179072952270508,
"step": 2800,
"token_acc": 0.9119147316330415
},
{
"epoch": 0.6236223858932316,
"grad_norm": 1.6640625,
"learning_rate": 9.334367233357169e-05,
"loss": 0.3253290939331055,
"step": 2900,
"token_acc": 0.909042319059297
},
{
"epoch": 0.6451266060964465,
"grad_norm": 2.015625,
"learning_rate": 9.274076850134519e-05,
"loss": 0.31794769287109376,
"step": 3000,
"token_acc": 0.9108269303100317
},
{
"epoch": 0.6666308262996613,
"grad_norm": 1.9296875,
"learning_rate": 9.211385632111469e-05,
"loss": 0.3040708923339844,
"step": 3100,
"token_acc": 0.9163664726951773
},
{
"epoch": 0.6881350465028762,
"grad_norm": 2.1875,
"learning_rate": 9.146328794204169e-05,
"loss": 0.3194514083862305,
"step": 3200,
"token_acc": 0.909515801428353
},
{
"epoch": 0.7096392667060911,
"grad_norm": 1.8203125,
"learning_rate": 9.078942880144971e-05,
"loss": 0.2973049736022949,
"step": 3300,
"token_acc": 0.915813655305054
},
{
"epoch": 0.731143486909306,
"grad_norm": 1.296875,
"learning_rate": 9.009265741955071e-05,
"loss": 0.3128537368774414,
"step": 3400,
"token_acc": 0.9134898467577565
},
{
"epoch": 0.7526477071125208,
"grad_norm": 2.15625,
"learning_rate": 8.937336518682273e-05,
"loss": 0.2994493865966797,
"step": 3500,
"token_acc": 0.9148646310872762
},
{
"epoch": 0.7741519273157357,
"grad_norm": 1.59375,
"learning_rate": 8.863195614415803e-05,
"loss": 0.3013328552246094,
"step": 3600,
"token_acc": 0.9161106454316849
},
{
"epoch": 0.7956561475189506,
"grad_norm": 1.4375,
"learning_rate": 8.78688467559051e-05,
"loss": 0.292550048828125,
"step": 3700,
"token_acc": 0.9173435563980641
},
{
"epoch": 0.8171603677221655,
"grad_norm": 1.5390625,
"learning_rate": 8.708446567593241e-05,
"loss": 0.29124902725219726,
"step": 3800,
"token_acc": 0.9190288906089831
},
{
"epoch": 0.8386645879253803,
"grad_norm": 1.8515625,
"learning_rate": 8.627925350684482e-05,
"loss": 0.2943286514282227,
"step": 3900,
"token_acc": 0.9158156459753924
},
{
"epoch": 0.8601688081285952,
"grad_norm": 1.5,
"learning_rate": 8.54536625524883e-05,
"loss": 0.2970516014099121,
"step": 4000,
"token_acc": 0.9162451525880965
},
{
"epoch": 0.8816730283318102,
"grad_norm": 1.6875,
"learning_rate": 8.460815656388175e-05,
"loss": 0.2960320854187012,
"step": 4100,
"token_acc": 0.9170335429769392
},
{
"epoch": 0.903177248535025,
"grad_norm": 1.7734375,
"learning_rate": 8.374321047871874e-05,
"loss": 0.2804668426513672,
"step": 4200,
"token_acc": 0.9197268393248292
},
{
"epoch": 0.9246814687382399,
"grad_norm": 1.328125,
"learning_rate": 8.28593101545855e-05,
"loss": 0.2943642807006836,
"step": 4300,
"token_acc": 0.9148772033647546
},
{
"epoch": 0.9461856889414547,
"grad_norm": 2.03125,
"learning_rate": 8.195695209604492e-05,
"loss": 0.27788606643676755,
"step": 4400,
"token_acc": 0.9207105105432354
},
{
"epoch": 0.9676899091446697,
"grad_norm": 1.53125,
"learning_rate": 8.103664317574e-05,
"loss": 0.2929634475708008,
"step": 4500,
"token_acc": 0.9166524756268892
},
{
"epoch": 0.9891941293478845,
"grad_norm": 1.3984375,
"learning_rate": 8.009890034967327e-05,
"loss": 0.2756942176818848,
"step": 4600,
"token_acc": 0.9201785658278533
},
{
"epoch": 1.0105370678995753,
"grad_norm": 2.15625,
"learning_rate": 7.914425036682224e-05,
"loss": 0.24342775344848633,
"step": 4700,
"token_acc": 0.9299042798987788
},
{
"epoch": 1.0320412881027903,
"grad_norm": 1.8671875,
"learning_rate": 7.817322947325397e-05,
"loss": 0.17995485305786132,
"step": 4800,
"token_acc": 0.9437043588941504
},
{
"epoch": 1.053545508306005,
"grad_norm": 1.4453125,
"learning_rate": 7.718638311090477e-05,
"loss": 0.180903377532959,
"step": 4900,
"token_acc": 0.9431556053516115
},
{
"epoch": 1.07504972850922,
"grad_norm": 1.4765625,
"learning_rate": 7.61842656111946e-05,
"loss": 0.18013723373413085,
"step": 5000,
"token_acc": 0.9432521310198607
},
{
"epoch": 1.0965539487124347,
"grad_norm": 1.46875,
"learning_rate": 7.516743988364796e-05,
"loss": 0.1733488655090332,
"step": 5100,
"token_acc": 0.9452036217975466
},
{
"epoch": 1.1180581689156497,
"grad_norm": 1.328125,
"learning_rate": 7.413647709969633e-05,
"loss": 0.18172847747802734,
"step": 5200,
"token_acc": 0.9450252493635491
},
{
"epoch": 1.1395623891188646,
"grad_norm": 1.7109375,
"learning_rate": 7.309195637183964e-05,
"loss": 0.1805746650695801,
"step": 5300,
"token_acc": 0.9436492682350668
},
{
"epoch": 1.1610666093220794,
"grad_norm": 1.6796875,
"learning_rate": 7.203446442834719e-05,
"loss": 0.17925085067749025,
"step": 5400,
"token_acc": 0.9439468707202018
},
{
"epoch": 1.1825708295252944,
"grad_norm": 1.4765625,
"learning_rate": 7.096459528368062e-05,
"loss": 0.169547119140625,
"step": 5500,
"token_acc": 0.9464933035229764
},
{
"epoch": 1.2040750497285093,
"grad_norm": 1.734375,
"learning_rate": 6.988294990482401e-05,
"loss": 0.17277034759521484,
"step": 5600,
"token_acc": 0.9460861364350833
},
{
"epoch": 1.225579269931724,
"grad_norm": 2.140625,
"learning_rate": 6.879013587370871e-05,
"loss": 0.17827749252319336,
"step": 5700,
"token_acc": 0.9442837392321599
},
{
"epoch": 1.247083490134939,
"grad_norm": 1.125,
"learning_rate": 6.768676704592228e-05,
"loss": 0.17549230575561522,
"step": 5800,
"token_acc": 0.9464748448974765
},
{
"epoch": 1.2685877103381538,
"grad_norm": 1.1953125,
"learning_rate": 6.657346320589358e-05,
"loss": 0.17796480178833007,
"step": 5900,
"token_acc": 0.9452043317064996
},
{
"epoch": 1.2900919305413687,
"grad_norm": 1.8359375,
"learning_rate": 6.545084971874738e-05,
"loss": 0.16906660079956054,
"step": 6000,
"token_acc": 0.9470721771309513
},
{
"epoch": 1.3115961507445837,
"grad_norm": 1.65625,
"learning_rate": 6.431955717902416e-05,
"loss": 0.16938840866088867,
"step": 6100,
"token_acc": 0.9465871890295269
},
{
"epoch": 1.3331003709477984,
"grad_norm": 1.6953125,
"learning_rate": 6.318022105646252e-05,
"loss": 0.16985233306884764,
"step": 6200,
"token_acc": 0.9453208613382422
},
{
"epoch": 1.3546045911510134,
"grad_norm": 1.421875,
"learning_rate": 6.203348133904315e-05,
"loss": 0.16554460525512696,
"step": 6300,
"token_acc": 0.9484946073875549
},
{
"epoch": 1.3761088113542281,
"grad_norm": 1.7265625,
"learning_rate": 6.087998217349456e-05,
"loss": 0.1714496612548828,
"step": 6400,
"token_acc": 0.94634329678104
},
{
"epoch": 1.397613031557443,
"grad_norm": 1.875,
"learning_rate": 5.972037150346305e-05,
"loss": 0.16627944946289064,
"step": 6500,
"token_acc": 0.9477827531417016
},
{
"epoch": 1.419117251760658,
"grad_norm": 1.453125,
"learning_rate": 5.855530070554975e-05,
"loss": 0.16439420700073243,
"step": 6600,
"token_acc": 0.9465121486955548
},
{
"epoch": 1.440621471963873,
"grad_norm": 1.0078125,
"learning_rate": 5.738542422341926e-05,
"loss": 0.16206270217895508,
"step": 6700,
"token_acc": 0.9495115245433108
},
{
"epoch": 1.4621256921670878,
"grad_norm": 1.625,
"learning_rate": 5.621139920018549e-05,
"loss": 0.1623272705078125,
"step": 6800,
"token_acc": 0.9498222630085545
},
{
"epoch": 1.4836299123703027,
"grad_norm": 1.546875,
"learning_rate": 5.5033885109281244e-05,
"loss": 0.16696163177490234,
"step": 6900,
"token_acc": 0.9472825094771207
},
{
"epoch": 1.5051341325735175,
"grad_norm": 1.578125,
"learning_rate": 5.3853543384018656e-05,
"loss": 0.16011791229248046,
"step": 7000,
"token_acc": 0.9493761372498051
},
{
"epoch": 1.5266383527767324,
"grad_norm": 1.2890625,
"learning_rate": 5.267103704604891e-05,
"loss": 0.16526294708251954,
"step": 7100,
"token_acc": 0.9476293902632834
},
{
"epoch": 1.5481425729799474,
"grad_norm": 1.828125,
"learning_rate": 5.148703033292958e-05,
"loss": 0.1586237907409668,
"step": 7200,
"token_acc": 0.9495203562756563
},
{
"epoch": 1.5696467931831624,
"grad_norm": 1.3671875,
"learning_rate": 5.030218832500914e-05,
"loss": 0.159231014251709,
"step": 7300,
"token_acc": 0.9495050052541342
},
{
"epoch": 1.591151013386377,
"grad_norm": 1.1171875,
"learning_rate": 4.911717657183785e-05,
"loss": 0.15366475105285646,
"step": 7400,
"token_acc": 0.9513064514402544
},
{
"epoch": 1.6126552335895918,
"grad_norm": 2.21875,
"learning_rate": 4.7932660718315374e-05,
"loss": 0.1546556568145752,
"step": 7500,
"token_acc": 0.9507159117503138
},
{
"epoch": 1.6341594537928068,
"grad_norm": 1.3984375,
"learning_rate": 4.6749306130784536e-05,
"loss": 0.1541473388671875,
"step": 7600,
"token_acc": 0.9510321351613004
},
{
"epoch": 1.6556636739960218,
"grad_norm": 1.390625,
"learning_rate": 4.556777752328174e-05,
"loss": 0.15245358467102052,
"step": 7700,
"token_acc": 0.9505623938879457
},
{
"epoch": 1.6771678941992367,
"grad_norm": 1.359375,
"learning_rate": 4.4388738584153746e-05,
"loss": 0.1586322021484375,
"step": 7800,
"token_acc": 0.9495154185022027
},
{
"epoch": 1.6986721144024515,
"grad_norm": 1.4375,
"learning_rate": 4.3212851603250584e-05,
"loss": 0.14727062225341797,
"step": 7900,
"token_acc": 0.953377205906427
},
{
"epoch": 1.7201763346056662,
"grad_norm": 1.1484375,
"learning_rate": 4.2040777099904036e-05,
"loss": 0.14960084915161131,
"step": 8000,
"token_acc": 0.9518475573738213
},
{
"epoch": 1.7416805548088812,
"grad_norm": 1.4765625,
"learning_rate": 4.087317345190068e-05,
"loss": 0.14496273040771485,
"step": 8100,
"token_acc": 0.9537553603397311
},
{
"epoch": 1.7631847750120961,
"grad_norm": 1.2265625,
"learning_rate": 3.9710696525657895e-05,
"loss": 0.1413516902923584,
"step": 8200,
"token_acc": 0.9545205949534054
},
{
"epoch": 1.784688995215311,
"grad_norm": 1.4296875,
"learning_rate": 3.855399930781046e-05,
"loss": 0.15235366821289062,
"step": 8300,
"token_acc": 0.9515693224431089
},
{
"epoch": 1.8061932154185258,
"grad_norm": 1.6015625,
"learning_rate": 3.740373153841489e-05,
"loss": 0.14871362686157227,
"step": 8400,
"token_acc": 0.9517884587043938
},
{
"epoch": 1.8276974356217408,
"grad_norm": 1.2265625,
"learning_rate": 3.626053934597737e-05,
"loss": 0.15128151893615724,
"step": 8500,
"token_acc": 0.951846881390593
},
{
"epoch": 1.8492016558249555,
"grad_norm": 1.4921875,
"learning_rate": 3.512506488451044e-05,
"loss": 0.14497539520263672,
"step": 8600,
"token_acc": 0.9535891224164648
},
{
"epoch": 1.8707058760281705,
"grad_norm": 1.3125,
"learning_rate": 3.399794597282213e-05,
"loss": 0.14073530197143555,
"step": 8700,
"token_acc": 0.9544612238812328
},
{
"epoch": 1.8922100962313855,
"grad_norm": 1.8359375,
"learning_rate": 3.287981573624039e-05,
"loss": 0.14946557998657226,
"step": 8800,
"token_acc": 0.9516438041229213
},
{
"epoch": 1.9137143164346004,
"grad_norm": 1.6640625,
"learning_rate": 3.177130225097384e-05,
"loss": 0.14700041770935057,
"step": 8900,
"token_acc": 0.95218424379823
},
{
"epoch": 1.9352185366378152,
"grad_norm": 1.3046875,
"learning_rate": 3.067302819130874e-05,
"loss": 0.14165225982666016,
"step": 9000,
"token_acc": 0.9545654059030715
},
{
"epoch": 1.95672275684103,
"grad_norm": 1.3359375,
"learning_rate": 2.9585610479840374e-05,
"loss": 0.14729736328125,
"step": 9100,
"token_acc": 0.9531239897411582
},
{
"epoch": 1.9782269770442449,
"grad_norm": 1.8984375,
"learning_rate": 2.850965994093523e-05,
"loss": 0.14732083320617675,
"step": 9200,
"token_acc": 0.9525314765967898
},
{
"epoch": 1.9997311972474598,
"grad_norm": 1.4453125,
"learning_rate": 2.744578095761868e-05,
"loss": 0.13918962478637695,
"step": 9300,
"token_acc": 0.9544093355357809
},
{
"epoch": 2.0210741357991506,
"grad_norm": 1.015625,
"learning_rate": 2.639457113208086e-05,
"loss": 0.052382316589355465,
"step": 9400,
"token_acc": 0.9839434949048363
},
{
"epoch": 2.0425783560023656,
"grad_norm": 0.98828125,
"learning_rate": 2.5356620949991606e-05,
"loss": 0.04628158092498779,
"step": 9500,
"token_acc": 0.9855887301570714
},
{
"epoch": 2.0640825762055806,
"grad_norm": 1.234375,
"learning_rate": 2.4332513448812706e-05,
"loss": 0.04719796180725098,
"step": 9600,
"token_acc": 0.9844260236121577
},
{
"epoch": 2.085586796408795,
"grad_norm": 1.4375,
"learning_rate": 2.3322823890294084e-05,
"loss": 0.045996718406677246,
"step": 9700,
"token_acc": 0.9853374587026064
},
{
"epoch": 2.10709101661201,
"grad_norm": 0.86328125,
"learning_rate": 2.2328119437337774e-05,
"loss": 0.04895151138305664,
"step": 9800,
"token_acc": 0.9847560313565394
},
{
"epoch": 2.128595236815225,
"grad_norm": 1.34375,
"learning_rate": 2.1348958835411008e-05,
"loss": 0.04724515438079834,
"step": 9900,
"token_acc": 0.9846966313539142
},
{
"epoch": 2.15009945701844,
"grad_norm": 1.2578125,
"learning_rate": 2.0385892098687814e-05,
"loss": 0.04533194541931152,
"step": 10000,
"token_acc": 0.9855662029433199
},
{
"epoch": 2.171603677221655,
"grad_norm": 1.4921875,
"learning_rate": 1.9439460201094854e-05,
"loss": 0.04482400417327881,
"step": 10100,
"token_acc": 0.9852925480820544
},
{
"epoch": 2.1931078974248694,
"grad_norm": 1.15625,
"learning_rate": 1.8510194772435658e-05,
"loss": 0.044654135704040525,
"step": 10200,
"token_acc": 0.9855592169025005
},
{
"epoch": 2.2146121176280844,
"grad_norm": 1.1171875,
"learning_rate": 1.7598617799763333e-05,
"loss": 0.044921741485595704,
"step": 10300,
"token_acc": 0.9858118111531502
},
{
"epoch": 2.2361163378312994,
"grad_norm": 1.40625,
"learning_rate": 1.670524133417013e-05,
"loss": 0.045565009117126465,
"step": 10400,
"token_acc": 0.9854009085317824
},
{
"epoch": 2.2576205580345143,
"grad_norm": 1.171875,
"learning_rate": 1.583056720315784e-05,
"loss": 0.04394767761230469,
"step": 10500,
"token_acc": 0.9859472514319911
},
{
"epoch": 2.2791247782377293,
"grad_norm": 1.875,
"learning_rate": 1.4975086728751459e-05,
"loss": 0.0461022424697876,
"step": 10600,
"token_acc": 0.9851400839279155
},
{
"epoch": 2.300628998440944,
"grad_norm": 1.296875,
"learning_rate": 1.4139280451513508e-05,
"loss": 0.045181169509887695,
"step": 10700,
"token_acc": 0.9857515277528576
},
{
"epoch": 2.322133218644159,
"grad_norm": 1.875,
"learning_rate": 1.332361786061494e-05,
"loss": 0.04347699165344238,
"step": 10800,
"token_acc": 0.9855772720203224
},
{
"epoch": 2.3436374388473737,
"grad_norm": 1.1953125,
"learning_rate": 1.2528557130113483e-05,
"loss": 0.045889606475830076,
"step": 10900,
"token_acc": 0.9852655048746884
},
{
"epoch": 2.3651416590505887,
"grad_norm": 0.98046875,
"learning_rate": 1.1754544861588207e-05,
"loss": 0.045331325531005856,
"step": 11000,
"token_acc": 0.9854636972109825
},
{
"epoch": 2.3866458792538037,
"grad_norm": 1.6015625,
"learning_rate": 1.1002015833274337e-05,
"loss": 0.043533525466918944,
"step": 11100,
"token_acc": 0.9858369327574255
},
{
"epoch": 2.4081500994570186,
"grad_norm": 1.515625,
"learning_rate": 1.0271392755839699e-05,
"loss": 0.0444044828414917,
"step": 11200,
"token_acc": 0.9853882193213813
},
{
"epoch": 2.429654319660233,
"grad_norm": 0.96484375,
"learning_rate": 9.563086034939484e-06,
"loss": 0.044610238075256346,
"step": 11300,
"token_acc": 0.9854142164675365
},
{
"epoch": 2.451158539863448,
"grad_norm": 1.1171875,
"learning_rate": 8.877493540683368e-06,
"loss": 0.043998885154724124,
"step": 11400,
"token_acc": 0.9856164758722682
},
{
"epoch": 2.472662760066663,
"grad_norm": 0.77734375,
"learning_rate": 8.215000384143674e-06,
"loss": 0.04451916217803955,
"step": 11500,
"token_acc": 0.9857173401753261
},
{
"epoch": 2.494166980269878,
"grad_norm": 1.4453125,
"learning_rate": 7.575978701030905e-06,
"loss": 0.04564218044281006,
"step": 11600,
"token_acc": 0.985550530171997
},
{
"epoch": 2.515671200473093,
"grad_norm": 1.3828125,
"learning_rate": 6.960787442657524e-06,
"loss": 0.04552034854888916,
"step": 11700,
"token_acc": 0.9857974537777493
},
{
"epoch": 2.5371754206763075,
"grad_norm": 1.5,
"learning_rate": 6.369772174307898e-06,
"loss": 0.0429258394241333,
"step": 11800,
"token_acc": 0.9856699458151402
},
{
"epoch": 2.5586796408795225,
"grad_norm": 1.3828125,
"learning_rate": 5.803264881127229e-06,
"loss": 0.04383339405059814,
"step": 11900,
"token_acc": 0.9856585097607918
},
{
"epoch": 2.5801838610827375,
"grad_norm": 0.96875,
"learning_rate": 5.26158378163894e-06,
"loss": 0.045687408447265626,
"step": 12000,
"token_acc": 0.9857526372659322
},
{
"epoch": 2.6016880812859524,
"grad_norm": 0.8046875,
"learning_rate": 4.7450331489948595e-06,
"loss": 0.045373907089233396,
"step": 12100,
"token_acc": 0.9854963714129514
},
{
"epoch": 2.6231923014891674,
"grad_norm": 1.28125,
"learning_rate": 4.253903140058907e-06,
"loss": 0.04461602687835693,
"step": 12200,
"token_acc": 0.9852423926254988
},
{
"epoch": 2.6446965216923823,
"grad_norm": 1.1015625,
"learning_rate": 3.7884696324201772e-06,
"loss": 0.043819246292114256,
"step": 12300,
"token_acc": 0.986441358611798
},
{
"epoch": 2.666200741895597,
"grad_norm": 1.2734375,
"learning_rate": 3.348994069426914e-06,
"loss": 0.0426084041595459,
"step": 12400,
"token_acc": 0.9861481886505206
},
{
"epoch": 2.687704962098812,
"grad_norm": 1.4140625,
"learning_rate": 2.935723313328587e-06,
"loss": 0.04416153907775879,
"step": 12500,
"token_acc": 0.9864983511071138
},
{
"epoch": 2.709209182302027,
"grad_norm": 1.1484375,
"learning_rate": 2.5488895066083485e-06,
"loss": 0.0465748405456543,
"step": 12600,
"token_acc": 0.9856129707559866
},
{
"epoch": 2.7307134025052417,
"grad_norm": 1.421875,
"learning_rate": 2.188709941583994e-06,
"loss": 0.04476622581481934,
"step": 12700,
"token_acc": 0.986246306274987
},
{
"epoch": 2.7522176227084563,
"grad_norm": 1.8359375,
"learning_rate": 1.855386938350462e-06,
"loss": 0.044062623977661135,
"step": 12800,
"token_acc": 0.9858560179250466
},
{
"epoch": 2.7737218429116712,
"grad_norm": 2.21875,
"learning_rate": 1.5491077311325764e-06,
"loss": 0.04457646369934082,
"step": 12900,
"token_acc": 0.9861132879847027
},
{
"epoch": 2.795226063114886,
"grad_norm": 1.21875,
"learning_rate": 1.2700443631118609e-06,
"loss": 0.04175244331359863,
"step": 13000,
"token_acc": 0.9865129911629239
},
{
"epoch": 2.816730283318101,
"grad_norm": 1.4921875,
"learning_rate": 1.0183535897863972e-06,
"loss": 0.044061436653137206,
"step": 13100,
"token_acc": 0.9862232779097387
},
{
"epoch": 2.838234503521316,
"grad_norm": 1.2421875,
"learning_rate": 7.941767909181053e-07,
"loss": 0.04133213520050049,
"step": 13200,
"token_acc": 0.9868434957201733
},
{
"epoch": 2.859738723724531,
"grad_norm": 1.0546875,
"learning_rate": 5.97639891116919e-07,
"loss": 0.04463282108306885,
"step": 13300,
"token_acc": 0.9855094470595891
},
{
"epoch": 2.881242943927746,
"grad_norm": 1.015625,
"learning_rate": 4.288532891063901e-07,
"loss": 0.040103521347045895,
"step": 13400,
"token_acc": 0.9871428724234936
},
{
"epoch": 2.9027471641309606,
"grad_norm": 1.4921875,
"learning_rate": 2.879117957105049e-07,
"loss": 0.04607905864715576,
"step": 13500,
"token_acc": 0.9850233112639176
},
{
"epoch": 2.9242513843341755,
"grad_norm": 1.234375,
"learning_rate": 1.7489458059654073e-07,
"loss": 0.044112420082092284,
"step": 13600,
"token_acc": 0.9855080248372473
},
{
"epoch": 2.9457556045373905,
"grad_norm": 1.0546875,
"learning_rate": 8.986512780389533e-08,
"loss": 0.04211245536804199,
"step": 13700,
"token_acc": 0.9865898745538862
},
{
"epoch": 2.9672598247406055,
"grad_norm": 1.0703125,
"learning_rate": 3.287120008378386e-08,
"loss": 0.0446782112121582,
"step": 13800,
"token_acc": 0.9859737628098912
},
{
"epoch": 2.98876404494382,
"grad_norm": 1.2578125,
"learning_rate": 3.94481206998365e-09,
"loss": 0.043461389541625976,
"step": 13900,
"token_acc": 0.9860611657341518
}
],
"logging_steps": 100,
"max_steps": 13953,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.979075114002413e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}