yoriis/multitask_model
011
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 1.0,6 "eval_steps": 500,7 "global_step": 5243,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.009536524890329964,14 "grad_norm": 1.3725634813308716,15 "learning_rate": 3.7262357414448674e-05,16 "loss": 1.5183903503417968,17 "step": 5018 },19 {20 "epoch": 0.019073049780659927,21 "grad_norm": 1.5630295276641846,22 "learning_rate": 7.52851711026616e-05,23 "loss": 1.0358538818359375,24 "step": 10025 },26 {27 "epoch": 0.02860957467098989,28 "grad_norm": 1.0457429885864258,29 "learning_rate": 0.00011330798479087452,30 "loss": 0.9928044128417969,31 "step": 15032 },33 {34 "epoch": 0.038146099561319854,35 "grad_norm": 1.1760847568511963,36 "learning_rate": 0.00015133079847908746,37 "loss": 0.9799901580810547,38 "step": 20039 },40 {41 "epoch": 0.04768262445164982,42 "grad_norm": 1.0198277235031128,43 "learning_rate": 0.0001893536121673004,44 "loss": 0.9692832946777343,45 "step": 25046 },47 {48 "epoch": 0.05721914934197978,49 "grad_norm": 4.2576141357421875,50 "learning_rate": 0.00019997421320309795,51 "loss": 0.9611747741699219,52 "step": 30053 },54 {55 "epoch": 0.06675567423230974,56 "grad_norm": 1.0672153234481812,57 "learning_rate": 0.00019985286992997873,58 "loss": 0.9739149475097656,59 "step": 35060 },61 {62 "epoch": 0.07629219912263971,63 "grad_norm": 1.015286922454834,64 "learning_rate": 0.00019963219089815488,65 "loss": 0.95861572265625,66 "step": 40067 },68 {69 "epoch": 0.08582872401296968,70 "grad_norm": 1.108496069908142,71 "learning_rate": 0.00019931239564382073,72 "loss": 0.9668975067138672,73 "step": 45074 },75 {76 "epoch": 0.09536524890329964,77 "grad_norm": 0.855691134929657,78 "learning_rate": 0.0001988938023060968,79 "loss": 0.9651553344726562,80 "step": 50081 },82 {83 "epoch": 0.09536524890329964,84 "eval_loss": 0.9762284159660339,85 "eval_runtime": 680.8217,86 "eval_samples_per_second": 12.536,87 "eval_steps_per_second": 0.784,88 "step": 50089 },90 {91 "epoch": 0.1049017737936296,92 "grad_norm": 0.8507081866264343,93 "learning_rate": 0.0001983768273105382,94 "loss": 0.9656350708007813,95 "step": 55096 },97 {98 "epoch": 0.11443829868395956,99 "grad_norm": 0.8325157165527344,100 "learning_rate": 0.00019776198495486565,101 "loss": 0.9318672180175781,102 "step": 600103 },104 {105 "epoch": 0.12397482357428953,106 "grad_norm": 0.9234330654144287,107 "learning_rate": 0.0001970498868973313,108 "loss": 0.9393777465820312,109 "step": 650110 },111 {112 "epoch": 0.13351134846461948,113 "grad_norm": 0.8355016708374023,114 "learning_rate": 0.0001962412415482278,115 "loss": 0.9270508575439453,116 "step": 700117 },118 {119 "epoch": 0.14304787335494945,120 "grad_norm": 0.9328432083129883,121 "learning_rate": 0.00019533685336514697,122 "loss": 0.9209387969970703,123 "step": 750124 },125 {126 "epoch": 0.15258439824527942,127 "grad_norm": 1.194154977798462,128 "learning_rate": 0.00019433762205268805,129 "loss": 0.9753380584716796,130 "step": 800131 },132 {133 "epoch": 0.16212092313560938,134 "grad_norm": 0.9554775953292847,135 "learning_rate": 0.0001932445416674127,136 "loss": 0.9683086395263671,137 "step": 850138 },139 {140 "epoch": 0.17165744802593935,141 "grad_norm": 1.0252937078475952,142 "learning_rate": 0.00019205869962893605,143 "loss": 0.9329315948486329,144 "step": 900145 },146 {147 "epoch": 0.18119397291626932,148 "grad_norm": 1.2285805940628052,149 "learning_rate": 0.00019078127563813883,150 "loss": 0.9511469268798828,151 "step": 950152 },153 {154 "epoch": 0.1907304978065993,155 "grad_norm": 0.9454661011695862,156 "learning_rate": 0.00018941354050357566,157 "loss": 0.951951904296875,158 "step": 1000159 },160 {161 "epoch": 0.1907304978065993,162 "eval_loss": 0.9598119258880615,163 "eval_runtime": 680.8627,164 "eval_samples_per_second": 12.536,165 "eval_steps_per_second": 0.784,166 "step": 1000167 },168 {169 "epoch": 0.20026702269692923,170 "grad_norm": 1.0210950374603271,171 "learning_rate": 0.00018795685487724782,172 "loss": 0.9263379669189453,173 "step": 1050174 },175 {176 "epoch": 0.2098035475872592,177 "grad_norm": 1.0087344646453857,178 "learning_rate": 0.0001864126679009975,179 "loss": 0.9310942840576172,180 "step": 1100181 },182 {183 "epoch": 0.21934007247758916,184 "grad_norm": 1.1398996114730835,185 "learning_rate": 0.00018478251576487092,186 "loss": 0.9465924072265625,187 "step": 1150188 },189 {190 "epoch": 0.22887659736791913,191 "grad_norm": 0.9276406764984131,192 "learning_rate": 0.0001830680201788836,193 "loss": 0.9552659606933593,194 "step": 1200195 },196 {197 "epoch": 0.2384131222582491,198 "grad_norm": 1.0002596378326416,199 "learning_rate": 0.00018127088675970888,200 "loss": 0.9471788024902343,201 "step": 1250202 },203 {204 "epoch": 0.24794964714857906,205 "grad_norm": 1.0108773708343506,206 "learning_rate": 0.000179392903333894,207 "loss": 0.9245248413085938,208 "step": 1300209 },210 {211 "epoch": 0.257486172038909,212 "grad_norm": 0.9765293002128601,213 "learning_rate": 0.0001774359381592925,214 "loss": 0.9204165649414062,215 "step": 1350216 },217 {218 "epoch": 0.26702269692923897,219 "grad_norm": 1.0059659481048584,220 "learning_rate": 0.00017540193806648134,221 "loss": 0.9452506256103516,222 "step": 1400223 },224 {225 "epoch": 0.27655922181956893,226 "grad_norm": 1.1858373880386353,227 "learning_rate": 0.0001732929265220125,228 "loss": 0.9286368560791015,229 "step": 1450230 },231 {232 "epoch": 0.2860957467098989,233 "grad_norm": 1.0006766319274902,234 "learning_rate": 0.00017111100161542545,235 "loss": 0.9553109741210938,236 "step": 1500237 },238 {239 "epoch": 0.2860957467098989,240 "eval_loss": 0.9582533836364746,241 "eval_runtime": 680.822,242 "eval_samples_per_second": 12.536,243 "eval_steps_per_second": 0.784,244 "step": 1500245 },246 {247 "epoch": 0.29563227160022887,248 "grad_norm": 0.9042516350746155,249 "learning_rate": 0.00016885833397202308,250 "loss": 0.93341796875,251 "step": 1550252 },253 {254 "epoch": 0.30516879649055884,255 "grad_norm": 0.9387173652648926,256 "learning_rate": 0.00016653716459348735,257 "loss": 0.9339485168457031,258 "step": 1600259 },260 {261 "epoch": 0.3147053213808888,262 "grad_norm": 0.9056106209754944,263 "learning_rate": 0.00016414980262848333,264 "loss": 0.9324442291259766,265 "step": 1650266 },267 {268 "epoch": 0.32424184627121877,269 "grad_norm": 0.9407602548599243,270 "learning_rate": 0.0001616986230754689,271 "loss": 0.9319153594970703,272 "step": 1700273 },274 {275 "epoch": 0.33377837116154874,276 "grad_norm": 0.9982315897941589,277 "learning_rate": 0.0001591860644199957,278 "loss": 0.9230084228515625,279 "step": 1750280 },281 {282 "epoch": 0.3433148960518787,283 "grad_norm": 1.0833745002746582,284 "learning_rate": 0.00015661462620885199,285 "loss": 0.9161262512207031,286 "step": 1800287 },288 {289 "epoch": 0.35285142094220867,290 "grad_norm": 0.9810793995857239,291 "learning_rate": 0.00015398686656346028,292 "loss": 0.9208243560791015,293 "step": 1850294 },295 {296 "epoch": 0.36238794583253864,297 "grad_norm": 1.094853401184082,298 "learning_rate": 0.00015130539963500376,299 "loss": 0.9387385559082031,300 "step": 1900301 },302 {303 "epoch": 0.3719244707228686,304 "grad_norm": 0.8733468055725098,305 "learning_rate": 0.000148572893003813,306 "loss": 0.9468997955322266,307 "step": 1950308 },309 {310 "epoch": 0.3814609956131986,311 "grad_norm": 1.0074750185012817,312 "learning_rate": 0.0001457920650256004,313 "loss": 0.9178851318359375,314 "step": 2000315 },316 {317 "epoch": 0.3814609956131986,318 "eval_loss": 0.9489485025405884,319 "eval_runtime": 681.5678,320 "eval_samples_per_second": 12.523,321 "eval_steps_per_second": 0.783,322 "step": 2000323 },324 {325 "epoch": 0.39099752050352854,326 "grad_norm": 1.0041425228118896,327 "learning_rate": 0.00014296568212718213,328 "loss": 0.9244281005859375,329 "step": 2050330 },331 {332 "epoch": 0.40053404539385845,333 "grad_norm": 1.0447967052459717,334 "learning_rate": 0.0001400965560543778,335 "loss": 0.9248933410644531,336 "step": 2100337 },338 {339 "epoch": 0.4100705702841884,340 "grad_norm": 0.968345046043396,341 "learning_rate": 0.00013718754107482596,342 "loss": 0.9109151458740234,343 "step": 2150344 },345 {346 "epoch": 0.4196070951745184,347 "grad_norm": 0.9737382531166077,348 "learning_rate": 0.0001342415311384981,349 "loss": 0.9231878662109375,350 "step": 2200351 },352 {353 "epoch": 0.42914362006484835,354 "grad_norm": 1.0494582653045654,355 "learning_rate": 0.00013126145699873532,356 "loss": 0.9193107604980468,357 "step": 2250358 },359 {360 "epoch": 0.4386801449551783,361 "grad_norm": 0.925656259059906,362 "learning_rate": 0.000128250283296673,363 "loss": 0.9101874542236328,364 "step": 2300365 },366 {367 "epoch": 0.4482166698455083,368 "grad_norm": 1.0119520425796509,369 "learning_rate": 0.00012521100561195233,370 "loss": 0.8976884460449219,371 "step": 2350372 },373 {374 "epoch": 0.45775319473583825,375 "grad_norm": 0.9749938249588013,376 "learning_rate": 0.0001221466474826543,377 "loss": 0.9223648834228516,378 "step": 2400379 },380 {381 "epoch": 0.4672897196261682,382 "grad_norm": 1.071219563484192,383 "learning_rate": 0.00011906025739741956,384 "loss": 0.9097858428955078,385 "step": 2450386 },387 {388 "epoch": 0.4768262445164982,389 "grad_norm": 1.0023361444473267,390 "learning_rate": 0.00011595490576274704,391 "loss": 0.9021361541748046,392 "step": 2500393 },394 {395 "epoch": 0.4768262445164982,396 "eval_loss": 0.9404354095458984,397 "eval_runtime": 681.2848,398 "eval_samples_per_second": 12.528,399 "eval_steps_per_second": 0.784,400 "step": 2500401 },402 {403 "epoch": 0.48636276940682815,404 "grad_norm": 0.7936119437217712,405 "learning_rate": 0.00011283368184848842,406 "loss": 0.9291069030761718,407 "step": 2550408 },409 {410 "epoch": 0.4958992942971581,411 "grad_norm": 0.8363732695579529,412 "learning_rate": 0.00010969969071457669,413 "loss": 0.8912797546386719,414 "step": 2600415 },416 {417 "epoch": 0.505435819187488,418 "grad_norm": 0.9047644734382629,419 "learning_rate": 0.0001065560501220464,420 "loss": 0.8831972503662109,421 "step": 2650422 },423 {424 "epoch": 0.514972344077818,425 "grad_norm": 0.8407217264175415,426 "learning_rate": 0.00010340588743141879,427 "loss": 0.9096057891845704,428 "step": 2700429 },430 {431 "epoch": 0.524508868968148,432 "grad_norm": 0.9929122924804688,433 "learning_rate": 0.00010025233649153707,434 "loss": 0.9299073028564453,435 "step": 2750436 },437 {438 "epoch": 0.5340453938584779,439 "grad_norm": 1.0364736318588257,440 "learning_rate": 9.70985345219468e-05,441 "loss": 0.9169498443603515,442 "step": 2800443 },444 {445 "epoch": 0.5435819187488079,446 "grad_norm": 0.9246217012405396,447 "learning_rate": 9.394761899192327e-05,448 "loss": 0.8965087127685547,449 "step": 2850450 },451 {452 "epoch": 0.5531184436391379,453 "grad_norm": 0.9324344992637634,454 "learning_rate": 9.08027244992503e-05,455 "loss": 0.9102237701416016,456 "step": 2900457 },458 {459 "epoch": 0.5626549685294678,460 "grad_norm": 0.998298704624176,461 "learning_rate": 8.766697965185565e-05,462 "loss": 0.9192097473144532,463 "step": 2950464 },465 {466 "epoch": 0.5721914934197978,467 "grad_norm": 0.9278152585029602,468 "learning_rate": 8.45435039554054e-05,469 "loss": 0.9202249908447265,470 "step": 3000471 },472 {473 "epoch": 0.5721914934197978,474 "eval_loss": 0.9327928423881531,475 "eval_runtime": 680.7868,476 "eval_samples_per_second": 12.537,477 "eval_steps_per_second": 0.784,478 "step": 3000479 },480 {481 "epoch": 0.5817280183101278,482 "grad_norm": 1.0032947063446045,483 "learning_rate": 8.14354047099533e-05,484 "loss": 0.9145193481445313,485 "step": 3050486 },487 {488 "epoch": 0.5912645432004577,489 "grad_norm": 0.834137499332428,490 "learning_rate": 7.834577391873266e-05,491 "loss": 0.9149667358398438,492 "step": 3100493 },494 {495 "epoch": 0.6008010680907877,496 "grad_norm": 0.9321780204772949,497 "learning_rate": 7.527768521216568e-05,498 "loss": 0.892784652709961,499 "step": 3150500 },501 {502 "epoch": 0.6103375929811177,503 "grad_norm": 1.0074846744537354,504 "learning_rate": 7.223419079015062e-05,505 "loss": 0.9072589111328125,506 "step": 3200507 },508 {509 "epoch": 0.6198741178714476,510 "grad_norm": 1.2590699195861816,511 "learning_rate": 6.921831838566842e-05,512 "loss": 0.9023927307128906,513 "step": 3250514 },515 {516 "epoch": 0.6294106427617776,517 "grad_norm": 0.7596737742424011,518 "learning_rate": 6.623306825272937e-05,519 "loss": 0.896091079711914,520 "step": 3300521 },522 {523 "epoch": 0.6389471676521076,524 "grad_norm": 0.9938393831253052,525 "learning_rate": 6.328141018165693e-05,526 "loss": 0.9024559783935547,527 "step": 3350528 },529 {530 "epoch": 0.6484836925424375,531 "grad_norm": 0.8195408582687378,532 "learning_rate": 6.036628054467682e-05,533 "loss": 0.8873242950439453,534 "step": 3400535 },536 {537 "epoch": 0.6580202174327675,538 "grad_norm": 0.9234633445739746,539 "learning_rate": 5.7490579374751686e-05,540 "loss": 0.8969253540039063,541 "step": 3450542 },543 {544 "epoch": 0.6675567423230975,545 "grad_norm": 0.9521735906600952,546 "learning_rate": 5.4657167480566594e-05,547 "loss": 0.8936698913574219,548 "step": 3500549 },550 {551 "epoch": 0.6675567423230975,552 "eval_loss": 0.9270365238189697,553 "eval_runtime": 680.6766,554 "eval_samples_per_second": 12.539,555 "eval_steps_per_second": 0.785,556 "step": 3500557 },558 {559 "epoch": 0.6770932672134274,560 "grad_norm": 0.9278208613395691,561 "learning_rate": 5.1868863600535646e-05,562 "loss": 0.8819551849365235,563 "step": 3550564 },565 {566 "epoch": 0.6866297921037574,567 "grad_norm": 0.8966768383979797,568 "learning_rate": 4.912844159866112e-05,569 "loss": 0.905844955444336,570 "step": 3600571 },572 {573 "epoch": 0.6961663169940874,574 "grad_norm": 0.886394202709198,575 "learning_rate": 4.6438627705034535e-05,576 "loss": 0.8963019561767578,577 "step": 3650578 },579 {580 "epoch": 0.7057028418844173,581 "grad_norm": 1.0538233518600464,582 "learning_rate": 4.380209780372496e-05,583 "loss": 0.8904299926757813,584 "step": 3700585 },586 {587 "epoch": 0.7152393667747473,588 "grad_norm": 1.0238487720489502,589 "learning_rate": 4.12214747707527e-05,590 "loss": 0.9049517822265625,591 "step": 3750592 },593 {594 "epoch": 0.7247758916650773,595 "grad_norm": 0.913487434387207,596 "learning_rate": 3.869932586479628e-05,597 "loss": 0.8899057006835938,598 "step": 3800599 },600 {601 "epoch": 0.7343124165554072,602 "grad_norm": 0.8777801990509033,603 "learning_rate": 3.623816017322917e-05,604 "loss": 0.8895880889892578,605 "step": 3850606 },607 {608 "epoch": 0.7438489414457372,609 "grad_norm": 1.0374138355255127,610 "learning_rate": 3.3840426116026044e-05,611 "loss": 0.9089337921142578,612 "step": 3900613 },614 {615 "epoch": 0.7533854663360672,616 "grad_norm": 1.0942944288253784,617 "learning_rate": 3.150850901002268e-05,618 "loss": 0.8874909210205079,619 "step": 3950620 },621 {622 "epoch": 0.7629219912263971,623 "grad_norm": 0.8818588256835938,624 "learning_rate": 2.9244728695951995e-05,625 "loss": 0.8881364440917969,626 "step": 4000627 },628 {629 "epoch": 0.7629219912263971,630 "eval_loss": 0.921961784362793,631 "eval_runtime": 680.8471,632 "eval_samples_per_second": 12.536,633 "eval_steps_per_second": 0.784,634 "step": 4000635 },636 {637 "epoch": 0.7724585161167271,638 "grad_norm": 0.8658433556556702,639 "learning_rate": 2.7051337230617125e-05,640 "loss": 0.887118148803711,641 "step": 4050642 },643 {644 "epoch": 0.7819950410070571,645 "grad_norm": 1.0360819101333618,646 "learning_rate": 2.4930516646497448e-05,647 "loss": 0.9029306030273437,648 "step": 4100649 },650 {651 "epoch": 0.791531565897387,652 "grad_norm": 0.9453035593032837,653 "learning_rate": 2.2884376781016258e-05,654 "loss": 0.874930419921875,655 "step": 4150656 },657 {658 "epoch": 0.8010680907877169,659 "grad_norm": 0.9671445488929749,660 "learning_rate": 2.0914953177629548e-05,661 "loss": 0.8965608978271484,662 "step": 4200663 },664 {665 "epoch": 0.8106046156780469,666 "grad_norm": 0.9346293210983276,667 "learning_rate": 1.902420506082424e-05,668 "loss": 0.8860881042480468,669 "step": 4250670 },671 {672 "epoch": 0.8201411405683768,673 "grad_norm": 1.0210144519805908,674 "learning_rate": 1.7214013387039884e-05,675 "loss": 0.9090773773193359,676 "step": 4300677 },678 {679 "epoch": 0.8296776654587068,680 "grad_norm": 0.7916406393051147,681 "learning_rate": 1.54861789734532e-05,682 "loss": 0.9077362060546875,683 "step": 4350684 },685 {686 "epoch": 0.8392141903490368,687 "grad_norm": 0.849738597869873,688 "learning_rate": 1.3842420706486903e-05,689 "loss": 0.8971955871582031,690 "step": 4400691 },692 {693 "epoch": 0.8487507152393667,694 "grad_norm": 0.8733137845993042,695 "learning_rate": 1.2284373831824847e-05,696 "loss": 0.8827657318115234,697 "step": 4450698 },699 {700 "epoch": 0.8582872401296967,701 "grad_norm": 0.9427047967910767,702 "learning_rate": 1.0813588327634961e-05,703 "loss": 0.8771236419677735,704 "step": 4500705 },706 {707 "epoch": 0.8582872401296967,708 "eval_loss": 0.9178871512413025,709 "eval_runtime": 680.8349,710 "eval_samples_per_second": 12.536,711 "eval_steps_per_second": 0.784,712 "step": 4500713 },714 {715 "epoch": 0.8678237650200267,716 "grad_norm": 0.8574934005737305,717 "learning_rate": 9.431527362617832e-06,718 "loss": 0.8978910827636719,719 "step": 4550720 },721 {722 "epoch": 0.8773602899103566,723 "grad_norm": 0.986375629901886,724 "learning_rate": 8.139565840415553e-06,725 "loss": 0.8829045867919922,726 "step": 4600727 },728 {729 "epoch": 0.8868968148006866,730 "grad_norm": 1.026666283607483,731 "learning_rate": 6.938989031828158e-06,732 "loss": 0.8786582183837891,733 "step": 4650734 },735 {736 "epoch": 0.8964333396910166,737 "grad_norm": 0.9898380041122437,738 "learning_rate": 5.8309912961990506e-06,739 "loss": 0.8815351867675781,740 "step": 4700741 },742 {743 "epoch": 0.9059698645813465,744 "grad_norm": 0.8167511224746704,745 "learning_rate": 4.8166748932408355e-06,746 "loss": 0.8737747192382812,747 "step": 4750748 },749 {750 "epoch": 0.9155063894716765,751 "grad_norm": 0.9592494368553162,752 "learning_rate": 3.8970488864839334e-06,753 "loss": 0.886360092163086,754 "step": 4800755 },756 {757 "epoch": 0.9250429143620065,758 "grad_norm": 1.1282484531402588,759 "learning_rate": 3.0730281394387382e-06,760 "loss": 0.9033355712890625,761 "step": 4850762 },763 {764 "epoch": 0.9345794392523364,765 "grad_norm": 1.0762990713119507,766 "learning_rate": 2.345432405469894e-06,767 "loss": 0.9018070983886719,768 "step": 4900769 },770 {771 "epoch": 0.9441159641426664,772 "grad_norm": 1.0391168594360352,773 "learning_rate": 1.7149855122882697e-06,774 "loss": 0.8773787689208984,775 "step": 4950776 },777 {778 "epoch": 0.9536524890329964,779 "grad_norm": 1.160536527633667,780 "learning_rate": 1.1823146418717068e-06,781 "loss": 0.8903836822509765,782 "step": 5000783 },784 {785 "epoch": 0.9536524890329964,786 "eval_loss": 0.9169394373893738,787 "eval_runtime": 681.071,788 "eval_samples_per_second": 12.532,789 "eval_steps_per_second": 0.784,790 "step": 5000791 },792 {793 "epoch": 0.9631890139233263,794 "grad_norm": 0.823947548866272,795 "learning_rate": 7.479497065310925e-07,796 "loss": 0.90203857421875,797 "step": 5050798 },799 {800 "epoch": 0.9727255388136563,801 "grad_norm": 0.7839481830596924,802 "learning_rate": 4.123228217422948e-07,803 "loss": 0.8997705078125,804 "step": 5100805 },806 {807 "epoch": 0.9822620637039863,808 "grad_norm": 0.9920501708984375,809 "learning_rate": 1.7576787626851777e-07,810 "loss": 0.8869709777832031,811 "step": 5150812 },813 {814 "epoch": 0.9917985885943162,815 "grad_norm": 0.9276648759841919,816 "learning_rate": 3.8520200000624615e-08,817 "loss": 0.8911021423339843,818 "step": 5200819 },820 {821 "epoch": 1.0,822 "eval_loss": 0.9169082045555115,823 "eval_runtime": 681.3749,824 "eval_samples_per_second": 12.526,825 "eval_steps_per_second": 0.784,826 "step": 5243827 }828 ],829 "logging_steps": 50,830 "max_steps": 5243,831 "num_input_tokens_seen": 0,832 "num_train_epochs": 1,833 "save_steps": 500,834 "stateful_callbacks": {835 "TrainerControl": {836 "args": {837 "should_epoch_stop": false,838 "should_evaluate": false,839 "should_log": false,840 "should_save": true,841 "should_training_stop": true842 },843 "attributes": {}844 }845 },846 "total_flos": 7.345861112836915e+18,847 "train_batch_size": 32,848 "trial_name": null,849 "trial_params": null850}851 