CoolFace
Modelpublic

Hartunka/tiny_bert_km_20_v1

sourceHugging Faceupdated 1y agoView on Hugging Face
0likes6downloads
trainer_state.json909 linesDownload Raw Back to root
1{2  "best_metric": 6.738729953765869,3  "best_model_checkpoint": "tiny_bert_km_20_v1/checkpoint-20000",4  "epoch": 25.0,5  "eval_steps": 10000,6  "global_step": 59550,7  "is_hyper_param_search": false,8  "is_local_process_zero": true,9  "is_world_process_zero": true,10  "log_history": [11    {12      "epoch": 0.20990764063811923,13      "grad_norm": 2.626917600631714,14      "learning_rate": 5e-06,15      "loss": 12.7573,16      "step": 50017    },18    {19      "epoch": 0.41981528127623846,20      "grad_norm": 1.9100428819656372,21      "learning_rate": 1e-05,22      "loss": 11.1676,23      "step": 100024    },25    {26      "epoch": 0.6297229219143576,27      "grad_norm": 2.924760341644287,28      "learning_rate": 1.5e-05,29      "loss": 9.7025,30      "step": 150031    },32    {33      "epoch": 0.8396305625524769,34      "grad_norm": 5.516970157623291,35      "learning_rate": 2e-05,36      "loss": 8.6156,37      "step": 200038    },39    {40      "epoch": 1.0495382031905962,41      "grad_norm": 8.952390670776367,42      "learning_rate": 2.5e-05,43      "loss": 7.9966,44      "step": 250045    },46    {47      "epoch": 1.2594458438287153,48      "grad_norm": 3.484361410140991,49      "learning_rate": 3e-05,50      "loss": 7.6978,51      "step": 300052    },53    {54      "epoch": 1.4693534844668346,55      "grad_norm": 7.40541934967041,56      "learning_rate": 3.5e-05,57      "loss": 7.5305,58      "step": 350059    },60    {61      "epoch": 1.6792611251049538,62      "grad_norm": 8.520888328552246,63      "learning_rate": 4e-05,64      "loss": 7.4397,65      "step": 400066    },67    {68      "epoch": 1.8891687657430731,69      "grad_norm": 6.013493061065674,70      "learning_rate": 4.5e-05,71      "loss": 7.3428,72      "step": 450073    },74    {75      "epoch": 2.0990764063811924,76      "grad_norm": 6.658591270446777,77      "learning_rate": 5e-05,78      "loss": 7.2303,79      "step": 500080    },81    {82      "epoch": 2.3089840470193117,83      "grad_norm": 3.478933334350586,84      "learning_rate": 5.500000000000001e-05,85      "loss": 7.1514,86      "step": 550087    },88    {89      "epoch": 2.5188916876574305,90      "grad_norm": 5.886797904968262,91      "learning_rate": 6e-05,92      "loss": 7.1208,93      "step": 600094    },95    {96      "epoch": 2.72879932829555,97      "grad_norm": 6.351716995239258,98      "learning_rate": 6.500000000000001e-05,99      "loss": 7.0765,100      "step": 6500101    },102    {103      "epoch": 2.938706968933669,104      "grad_norm": 9.912898063659668,105      "learning_rate": 7e-05,106      "loss": 7.0506,107      "step": 7000108    },109    {110      "epoch": 3.1486146095717884,111      "grad_norm": 5.107701778411865,112      "learning_rate": 7.500000000000001e-05,113      "loss": 6.9614,114      "step": 7500115    },116    {117      "epoch": 3.3585222502099077,118      "grad_norm": 5.066652297973633,119      "learning_rate": 8e-05,120      "loss": 6.9374,121      "step": 8000122    },123    {124      "epoch": 3.568429890848027,125      "grad_norm": 4.600908279418945,126      "learning_rate": 8.5e-05,127      "loss": 6.9234,128      "step": 8500129    },130    {131      "epoch": 3.7783375314861463,132      "grad_norm": 6.573398590087891,133      "learning_rate": 9e-05,134      "loss": 6.9026,135      "step": 9000136    },137    {138      "epoch": 3.988245172124265,139      "grad_norm": 4.131033897399902,140      "learning_rate": 9.5e-05,141      "loss": 6.8857,142      "step": 9500143    },144    {145      "epoch": 4.198152812762385,146      "grad_norm": 3.8546698093414307,147      "learning_rate": 0.0001,148      "loss": 6.7928,149      "step": 10000150    },151    {152      "epoch": 4.198152812762385,153      "eval_accuracy": 0.1481120788001892,154      "eval_loss": 6.862878799438477,155      "eval_runtime": 0.699,156      "eval_samples_per_second": 685.226,157      "eval_steps_per_second": 7.153,158      "step": 10000159    },160    {161      "epoch": 4.408060453400504,162      "grad_norm": 3.9628775119781494,163      "learning_rate": 9.899091826437942e-05,164      "loss": 6.7997,165      "step": 10500166    },167    {168      "epoch": 4.617968094038623,169      "grad_norm": 2.652864694595337,170      "learning_rate": 9.798183652875883e-05,171      "loss": 6.7921,172      "step": 11000173    },174    {175      "epoch": 4.827875734676742,176      "grad_norm": 2.7218310832977295,177      "learning_rate": 9.697275479313825e-05,178      "loss": 6.7781,179      "step": 11500180    },181    {182      "epoch": 5.037783375314861,183      "grad_norm": 3.650275945663452,184      "learning_rate": 9.596367305751767e-05,185      "loss": 6.7365,186      "step": 12000187    },188    {189      "epoch": 5.247691015952981,190      "grad_norm": 2.891012191772461,191      "learning_rate": 9.495459132189707e-05,192      "loss": 6.6725,193      "step": 12500194    },195    {196      "epoch": 5.4575986565911,197      "grad_norm": 3.2625622749328613,198      "learning_rate": 9.394550958627649e-05,199      "loss": 6.6693,200      "step": 13000201    },202    {203      "epoch": 5.667506297229219,204      "grad_norm": 2.9605414867401123,205      "learning_rate": 9.293642785065591e-05,206      "loss": 6.6694,207      "step": 13500208    },209    {210      "epoch": 5.877413937867338,211      "grad_norm": 2.8077549934387207,212      "learning_rate": 9.192734611503532e-05,213      "loss": 6.6607,214      "step": 14000215    },216    {217      "epoch": 6.087321578505458,218      "grad_norm": 4.0943121910095215,219      "learning_rate": 9.091826437941474e-05,220      "loss": 6.6173,221      "step": 14500222    },223    {224      "epoch": 6.297229219143577,225      "grad_norm": 2.0766336917877197,226      "learning_rate": 8.990918264379415e-05,227      "loss": 6.5676,228      "step": 15000229    },230    {231      "epoch": 6.507136859781696,232      "grad_norm": 2.148397445678711,233      "learning_rate": 8.890010090817356e-05,234      "loss": 6.5737,235      "step": 15500236    },237    {238      "epoch": 6.717044500419815,239      "grad_norm": 2.837857961654663,240      "learning_rate": 8.789101917255298e-05,241      "loss": 6.5691,242      "step": 16000243    },244    {245      "epoch": 6.926952141057934,246      "grad_norm": 2.902109384536743,247      "learning_rate": 8.68819374369324e-05,248      "loss": 6.5689,249      "step": 16500250    },251    {252      "epoch": 7.136859781696054,253      "grad_norm": 1.992867350578308,254      "learning_rate": 8.58728557013118e-05,255      "loss": 6.5103,256      "step": 17000257    },258    {259      "epoch": 7.346767422334173,260      "grad_norm": 4.094528675079346,261      "learning_rate": 8.486377396569122e-05,262      "loss": 6.4893,263      "step": 17500264    },265    {266      "epoch": 7.5566750629722925,267      "grad_norm": 3.531911849975586,268      "learning_rate": 8.385469223007064e-05,269      "loss": 6.4962,270      "step": 18000271    },272    {273      "epoch": 7.766582703610411,274      "grad_norm": 3.1731412410736084,275      "learning_rate": 8.284561049445006e-05,276      "loss": 6.4986,277      "step": 18500278    },279    {280      "epoch": 7.97649034424853,281      "grad_norm": 2.611910820007324,282      "learning_rate": 8.183652875882947e-05,283      "loss": 6.4964,284      "step": 19000285    },286    {287      "epoch": 8.18639798488665,288      "grad_norm": 2.186275005340576,289      "learning_rate": 8.082744702320889e-05,290      "loss": 6.4235,291      "step": 19500292    },293    {294      "epoch": 8.39630562552477,295      "grad_norm": 2.4784603118896484,296      "learning_rate": 7.98183652875883e-05,297      "loss": 6.4284,298      "step": 20000299    },300    {301      "epoch": 8.39630562552477,302      "eval_accuracy": 0.15283818288953113,303      "eval_loss": 6.738729953765869,304      "eval_runtime": 0.6762,305      "eval_samples_per_second": 708.384,306      "eval_steps_per_second": 7.394,307      "step": 20000308    },309    {310      "epoch": 8.606213266162888,311      "grad_norm": 3.508121967315674,312      "learning_rate": 7.880928355196771e-05,313      "loss": 6.4296,314      "step": 20500315    },316    {317      "epoch": 8.816120906801007,318      "grad_norm": 3.2439723014831543,319      "learning_rate": 7.780020181634713e-05,320      "loss": 6.4308,321      "step": 21000322    },323    {324      "epoch": 9.026028547439127,325      "grad_norm": 2.687647819519043,326      "learning_rate": 7.679112008072655e-05,327      "loss": 6.4231,328      "step": 21500329    },330    {331      "epoch": 9.235936188077247,332      "grad_norm": 2.5599114894866943,333      "learning_rate": 7.578203834510595e-05,334      "loss": 6.3691,335      "step": 22000336    },337    {338      "epoch": 9.445843828715365,339      "grad_norm": 2.2765402793884277,340      "learning_rate": 7.477295660948537e-05,341      "loss": 6.374,342      "step": 22500343    },344    {345      "epoch": 9.655751469353484,346      "grad_norm": 5.340300559997559,347      "learning_rate": 7.376387487386479e-05,348      "loss": 6.3733,349      "step": 23000350    },351    {352      "epoch": 9.865659109991604,353      "grad_norm": 3.3845908641815186,354      "learning_rate": 7.27547931382442e-05,355      "loss": 6.3774,356      "step": 23500357    },358    {359      "epoch": 10.075566750629722,360      "grad_norm": 2.933563709259033,361      "learning_rate": 7.174571140262362e-05,362      "loss": 6.3517,363      "step": 24000364    },365    {366      "epoch": 10.285474391267842,367      "grad_norm": 4.057340145111084,368      "learning_rate": 7.073662966700304e-05,369      "loss": 6.3229,370      "step": 24500371    },372    {373      "epoch": 10.495382031905962,374      "grad_norm": 2.4048781394958496,375      "learning_rate": 6.972754793138244e-05,376      "loss": 6.326,377      "step": 25000378    },379    {380      "epoch": 10.705289672544081,381      "grad_norm": 3.3523154258728027,382      "learning_rate": 6.871846619576186e-05,383      "loss": 6.3318,384      "step": 25500385    },386    {387      "epoch": 10.9151973131822,388      "grad_norm": 3.0950045585632324,389      "learning_rate": 6.770938446014128e-05,390      "loss": 6.3367,391      "step": 26000392    },393    {394      "epoch": 11.125104953820319,395      "grad_norm": 2.1324470043182373,396      "learning_rate": 6.670030272452068e-05,397      "loss": 6.2982,398      "step": 26500399    },400    {401      "epoch": 11.335012594458439,402      "grad_norm": 3.2120983600616455,403      "learning_rate": 6.56912209889001e-05,404      "loss": 6.2809,405      "step": 27000406    },407    {408      "epoch": 11.544920235096557,409      "grad_norm": 3.337730884552002,410      "learning_rate": 6.468213925327952e-05,411      "loss": 6.2798,412      "step": 27500413    },414    {415      "epoch": 11.754827875734676,416      "grad_norm": 3.8394360542297363,417      "learning_rate": 6.367305751765893e-05,418      "loss": 6.2886,419      "step": 28000420    },421    {422      "epoch": 11.964735516372796,423      "grad_norm": 1.6199002265930176,424      "learning_rate": 6.266397578203835e-05,425      "loss": 6.2853,426      "step": 28500427    },428    {429      "epoch": 12.174643157010916,430      "grad_norm": 3.1857144832611084,431      "learning_rate": 6.165489404641777e-05,432      "loss": 6.2519,433      "step": 29000434    },435    {436      "epoch": 12.384550797649034,437      "grad_norm": 1.888143539428711,438      "learning_rate": 6.064581231079718e-05,439      "loss": 6.2412,440      "step": 29500441    },442    {443      "epoch": 12.594458438287154,444      "grad_norm": 2.276371479034424,445      "learning_rate": 5.963673057517659e-05,446      "loss": 6.2528,447      "step": 30000448    },449    {450      "epoch": 12.594458438287154,451      "eval_accuracy": 0.15337045638973612,452      "eval_loss": 6.793837070465088,453      "eval_runtime": 0.6792,454      "eval_samples_per_second": 705.229,455      "eval_steps_per_second": 7.361,456      "step": 30000457    },458    {459      "epoch": 12.804366078925273,460      "grad_norm": 2.553093671798706,461      "learning_rate": 5.8627648839556e-05,462      "loss": 6.2482,463      "step": 30500464    },465    {466      "epoch": 13.014273719563391,467      "grad_norm": 2.7331161499023438,468      "learning_rate": 5.7618567103935416e-05,469      "loss": 6.2494,470      "step": 31000471    },472    {473      "epoch": 13.224181360201511,474      "grad_norm": 1.9387927055358887,475      "learning_rate": 5.660948536831484e-05,476      "loss": 6.2118,477      "step": 31500478    },479    {480      "epoch": 13.43408900083963,481      "grad_norm": 2.0029964447021484,482      "learning_rate": 5.5600403632694254e-05,483      "loss": 6.21,484      "step": 32000485    },486    {487      "epoch": 13.64399664147775,488      "grad_norm": 2.053971529006958,489      "learning_rate": 5.4591321897073666e-05,490      "loss": 6.2202,491      "step": 32500492    },493    {494      "epoch": 13.853904282115868,495      "grad_norm": 2.6166188716888428,496      "learning_rate": 5.358224016145308e-05,497      "loss": 6.22,498      "step": 33000499    },500    {501      "epoch": 14.063811922753988,502      "grad_norm": 3.489931106567383,503      "learning_rate": 5.257315842583249e-05,504      "loss": 6.2067,505      "step": 33500506    },507    {508      "epoch": 14.273719563392108,509      "grad_norm": 1.5623815059661865,510      "learning_rate": 5.15640766902119e-05,511      "loss": 6.1869,512      "step": 34000513    },514    {515      "epoch": 14.483627204030226,516      "grad_norm": 1.6946079730987549,517      "learning_rate": 5.055499495459133e-05,518      "loss": 6.1897,519      "step": 34500520    },521    {522      "epoch": 14.693534844668346,523      "grad_norm": 2.893101215362549,524      "learning_rate": 4.954591321897074e-05,525      "loss": 6.1905,526      "step": 35000527    },528    {529      "epoch": 14.903442485306465,530      "grad_norm": 3.0789566040039062,531      "learning_rate": 4.853683148335015e-05,532      "loss": 6.1918,533      "step": 35500534    },535    {536      "epoch": 15.113350125944585,537      "grad_norm": 2.1385488510131836,538      "learning_rate": 4.7527749747729566e-05,539      "loss": 6.1725,540      "step": 36000541    },542    {543      "epoch": 15.323257766582703,544      "grad_norm": 1.7395439147949219,545      "learning_rate": 4.6518668012108985e-05,546      "loss": 6.1673,547      "step": 36500548    },549    {550      "epoch": 15.533165407220823,551      "grad_norm": 3.8336212635040283,552      "learning_rate": 4.55095862764884e-05,553      "loss": 6.1665,554      "step": 37000555    },556    {557      "epoch": 15.743073047858942,558      "grad_norm": 2.4808847904205322,559      "learning_rate": 4.450050454086781e-05,560      "loss": 6.1688,561      "step": 37500562    },563    {564      "epoch": 15.95298068849706,565      "grad_norm": 3.9198453426361084,566      "learning_rate": 4.349142280524723e-05,567      "loss": 6.1669,568      "step": 38000569    },570    {571      "epoch": 16.162888329135182,572      "grad_norm": 3.0992448329925537,573      "learning_rate": 4.248234106962664e-05,574      "loss": 6.1483,575      "step": 38500576    },577    {578      "epoch": 16.3727959697733,579      "grad_norm": 1.923215389251709,580      "learning_rate": 4.147325933400605e-05,581      "loss": 6.1389,582      "step": 39000583    },584    {585      "epoch": 16.582703610411418,586      "grad_norm": 2.8576881885528564,587      "learning_rate": 4.046417759838547e-05,588      "loss": 6.1462,589      "step": 39500590    },591    {592      "epoch": 16.79261125104954,593      "grad_norm": 2.547161817550659,594      "learning_rate": 3.9455095862764884e-05,595      "loss": 6.135,596      "step": 40000597    },598    {599      "epoch": 16.79261125104954,600      "eval_accuracy": 0.15220926320511383,601      "eval_loss": 6.990812301635742,602      "eval_runtime": 0.6803,603      "eval_samples_per_second": 704.08,604      "eval_steps_per_second": 7.349,605      "step": 40000606    },607    {608      "epoch": 17.002518891687657,609      "grad_norm": 2.547924041748047,610      "learning_rate": 3.84460141271443e-05,611      "loss": 6.1427,612      "step": 40500613    },614    {615      "epoch": 17.212426532325775,616      "grad_norm": 1.9287402629852295,617      "learning_rate": 3.7436932391523715e-05,618      "loss": 6.119,619      "step": 41000620    },621    {622      "epoch": 17.422334172963897,623      "grad_norm": 3.7875235080718994,624      "learning_rate": 3.642785065590313e-05,625      "loss": 6.1214,626      "step": 41500627    },628    {629      "epoch": 17.632241813602015,630      "grad_norm": 2.060349941253662,631      "learning_rate": 3.541876892028255e-05,632      "loss": 6.1224,633      "step": 42000634    },635    {636      "epoch": 17.842149454240133,637      "grad_norm": 1.9006288051605225,638      "learning_rate": 3.440968718466196e-05,639      "loss": 6.1238,640      "step": 42500641    },642    {643      "epoch": 18.052057094878254,644      "grad_norm": 3.339731216430664,645      "learning_rate": 3.340060544904137e-05,646      "loss": 6.1152,647      "step": 43000648    },649    {650      "epoch": 18.261964735516372,651      "grad_norm": 3.107632637023926,652      "learning_rate": 3.239152371342079e-05,653      "loss": 6.1032,654      "step": 43500655    },656    {657      "epoch": 18.471872376154494,658      "grad_norm": 3.526423215866089,659      "learning_rate": 3.13824419778002e-05,660      "loss": 6.1028,661      "step": 44000662    },663    {664      "epoch": 18.68178001679261,665      "grad_norm": 2.2601630687713623,666      "learning_rate": 3.0373360242179615e-05,667      "loss": 6.1045,668      "step": 44500669    },670    {671      "epoch": 18.89168765743073,672      "grad_norm": 3.369821071624756,673      "learning_rate": 2.9364278506559034e-05,674      "loss": 6.1077,675      "step": 45000676    },677    {678      "epoch": 19.10159529806885,679      "grad_norm": 2.336292266845703,680      "learning_rate": 2.8355196770938446e-05,681      "loss": 6.0984,682      "step": 45500683    },684    {685      "epoch": 19.31150293870697,686      "grad_norm": 1.6681281328201294,687      "learning_rate": 2.734611503531786e-05,688      "loss": 6.0841,689      "step": 46000690    },691    {692      "epoch": 19.521410579345087,693      "grad_norm": 1.2452404499053955,694      "learning_rate": 2.6337033299697278e-05,695      "loss": 6.0884,696      "step": 46500697    },698    {699      "epoch": 19.73131821998321,700      "grad_norm": 2.2260518074035645,701      "learning_rate": 2.532795156407669e-05,702      "loss": 6.0906,703      "step": 47000704    },705    {706      "epoch": 19.941225860621326,707      "grad_norm": 2.4123737812042236,708      "learning_rate": 2.4318869828456106e-05,709      "loss": 6.0857,710      "step": 47500711    },712    {713      "epoch": 20.151133501259444,714      "grad_norm": 0.9671133160591125,715      "learning_rate": 2.330978809283552e-05,716      "loss": 6.0772,717      "step": 48000718    },719    {720      "epoch": 20.361041141897566,721      "grad_norm": 3.584923028945923,722      "learning_rate": 2.2300706357214934e-05,723      "loss": 6.076,724      "step": 48500725    },726    {727      "epoch": 20.570948782535684,728      "grad_norm": 2.9383702278137207,729      "learning_rate": 2.129162462159435e-05,730      "loss": 6.0682,731      "step": 49000732    },733    {734      "epoch": 20.7808564231738,735      "grad_norm": 2.90152645111084,736      "learning_rate": 2.0282542885973765e-05,737      "loss": 6.07,738      "step": 49500739    },740    {741      "epoch": 20.990764063811923,742      "grad_norm": 2.645386219024658,743      "learning_rate": 1.927346115035318e-05,744      "loss": 6.0725,745      "step": 50000746    },747    {748      "epoch": 20.990764063811923,749      "eval_accuracy": 0.1516247511705498,750      "eval_loss": 7.07234525680542,751      "eval_runtime": 0.7081,752      "eval_samples_per_second": 676.413,753      "eval_steps_per_second": 7.061,754      "step": 50000755    },756    {757      "epoch": 21.20067170445004,758      "grad_norm": 2.878293752670288,759      "learning_rate": 1.8264379414732593e-05,760      "loss": 6.0613,761      "step": 50500762    },763    {764      "epoch": 21.410579345088163,765      "grad_norm": 3.8497214317321777,766      "learning_rate": 1.725529767911201e-05,767      "loss": 6.0621,768      "step": 51000769    },770    {771      "epoch": 21.62048698572628,772      "grad_norm": 1.451300859451294,773      "learning_rate": 1.6246215943491424e-05,774      "loss": 6.0617,775      "step": 51500776    },777    {778      "epoch": 21.8303946263644,779      "grad_norm": 2.1480000019073486,780      "learning_rate": 1.5237134207870837e-05,781      "loss": 6.0614,782      "step": 52000783    },784    {785      "epoch": 22.04030226700252,786      "grad_norm": 1.1892906427383423,787      "learning_rate": 1.4228052472250252e-05,788      "loss": 6.0609,789      "step": 52500790    },791    {792      "epoch": 22.250209907640638,793      "grad_norm": 0.9119521379470825,794      "learning_rate": 1.3218970736629668e-05,795      "loss": 6.0453,796      "step": 53000797    },798    {799      "epoch": 22.460117548278756,800      "grad_norm": 1.1588727235794067,801      "learning_rate": 1.2209889001009082e-05,802      "loss": 6.0493,803      "step": 53500804    },805    {806      "epoch": 22.670025188916878,807      "grad_norm": 0.8290802240371704,808      "learning_rate": 1.1200807265388496e-05,809      "loss": 6.0531,810      "step": 54000811    },812    {813      "epoch": 22.879932829554996,814      "grad_norm": 2.3373863697052,815      "learning_rate": 1.0191725529767912e-05,816      "loss": 6.0472,817      "step": 54500818    },819    {820      "epoch": 23.089840470193113,821      "grad_norm": 1.9802433252334595,822      "learning_rate": 9.182643794147326e-06,823      "loss": 6.0441,824      "step": 55000825    },826    {827      "epoch": 23.299748110831235,828      "grad_norm": 0.8753450512886047,829      "learning_rate": 8.173562058526741e-06,830      "loss": 6.0393,831      "step": 55500832    },833    {834      "epoch": 23.509655751469353,835      "grad_norm": 2.932295560836792,836      "learning_rate": 7.164480322906156e-06,837      "loss": 6.0449,838      "step": 56000839    },840    {841      "epoch": 23.719563392107474,842      "grad_norm": 3.6671857833862305,843      "learning_rate": 6.155398587285571e-06,844      "loss": 6.0376,845      "step": 56500846    },847    {848      "epoch": 23.929471032745592,849      "grad_norm": 1.280537724494934,850      "learning_rate": 5.146316851664986e-06,851      "loss": 6.0395,852      "step": 57000853    },854    {855      "epoch": 24.13937867338371,856      "grad_norm": 1.0981942415237427,857      "learning_rate": 4.1372351160444e-06,858      "loss": 6.0356,859      "step": 57500860    },861    {862      "epoch": 24.349286314021832,863      "grad_norm": 2.460625410079956,864      "learning_rate": 3.1281533804238145e-06,865      "loss": 6.0385,866      "step": 58000867    },868    {869      "epoch": 24.55919395465995,870      "grad_norm": 2.348940372467041,871      "learning_rate": 2.1190716448032293e-06,872      "loss": 6.0341,873      "step": 58500874    },875    {876      "epoch": 24.769101595298068,877      "grad_norm": 1.964220404624939,878      "learning_rate": 1.109989909182644e-06,879      "loss": 6.0323,880      "step": 59000881    },882    {883      "epoch": 24.97900923593619,884      "grad_norm": 1.1357017755508423,885      "learning_rate": 1.0090817356205853e-07,886      "loss": 6.031,887      "step": 59500888    },889    {890      "epoch": 25.0,891      "step": 59550,892      "total_flos": 3.051148107142656e+17,893      "train_loss": 6.524913863046744,894      "train_runtime": 11619.3207,895      "train_samples_per_second": 491.937,896      "train_steps_per_second": 5.125897    }898  ],899  "logging_steps": 500,900  "max_steps": 59550,901  "num_input_tokens_seen": 0,902  "num_train_epochs": 25,903  "save_steps": 10000,904  "total_flos": 3.051148107142656e+17,905  "train_batch_size": 96,906  "trial_name": null,907  "trial_params": null908}909