Hartunka/bert_base_km_100_v2
014
1{2 "best_global_step": 20000,3 "best_metric": 7.036741733551025,4 "best_model_checkpoint": "bert_base_km_100_v2/checkpoint-20000",5 "epoch": 25.0,6 "eval_steps": 10000,7 "global_step": 59550,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.20990764063811923,14 "grad_norm": 3.0325520038604736,15 "learning_rate": 5e-06,16 "loss": 13.7954,17 "step": 50018 },19 {20 "epoch": 0.41981528127623846,21 "grad_norm": 2.8106133937835693,22 "learning_rate": 1e-05,23 "loss": 12.0341,24 "step": 100025 },26 {27 "epoch": 0.6297229219143576,28 "grad_norm": 11.055712699890137,29 "learning_rate": 1.5e-05,30 "loss": 10.5831,31 "step": 150032 },33 {34 "epoch": 0.8396305625524769,35 "grad_norm": 5.347803115844727,36 "learning_rate": 2e-05,37 "loss": 9.7271,38 "step": 200039 },40 {41 "epoch": 1.0495382031905962,42 "grad_norm": 15.038311004638672,43 "learning_rate": 2.5e-05,44 "loss": 9.1027,45 "step": 250046 },47 {48 "epoch": 1.2594458438287153,49 "grad_norm": 17.481904983520508,50 "learning_rate": 3e-05,51 "loss": 8.6071,52 "step": 300053 },54 {55 "epoch": 1.4693534844668346,56 "grad_norm": 14.806625366210938,57 "learning_rate": 3.5e-05,58 "loss": 8.2317,59 "step": 350060 },61 {62 "epoch": 1.6792611251049538,63 "grad_norm": 15.759472846984863,64 "learning_rate": 4e-05,65 "loss": 8.0202,66 "step": 400067 },68 {69 "epoch": 1.8891687657430731,70 "grad_norm": 7.746140480041504,71 "learning_rate": 4.5e-05,72 "loss": 7.8677,73 "step": 450074 },75 {76 "epoch": 2.0990764063811924,77 "grad_norm": 9.09663200378418,78 "learning_rate": 5e-05,79 "loss": 7.699,80 "step": 500081 },82 {83 "epoch": 2.3089840470193117,84 "grad_norm": 8.326930046081543,85 "learning_rate": 5.500000000000001e-05,86 "loss": 7.5839,87 "step": 550088 },89 {90 "epoch": 2.5188916876574305,91 "grad_norm": 4.7854905128479,92 "learning_rate": 6e-05,93 "loss": 7.5285,94 "step": 600095 },96 {97 "epoch": 2.72879932829555,98 "grad_norm": 5.352976322174072,99 "learning_rate": 6.500000000000001e-05,100 "loss": 7.5069,101 "step": 6500102 },103 {104 "epoch": 2.938706968933669,105 "grad_norm": 6.005526065826416,106 "learning_rate": 7e-05,107 "loss": 7.4692,108 "step": 7000109 },110 {111 "epoch": 3.1486146095717884,112 "grad_norm": 5.010321617126465,113 "learning_rate": 7.500000000000001e-05,114 "loss": 7.3121,115 "step": 7500116 },117 {118 "epoch": 3.3585222502099077,119 "grad_norm": 5.591887950897217,120 "learning_rate": 8e-05,121 "loss": 7.2788,122 "step": 8000123 },124 {125 "epoch": 3.568429890848027,126 "grad_norm": 4.3518385887146,127 "learning_rate": 8.5e-05,128 "loss": 7.2869,129 "step": 8500130 },131 {132 "epoch": 3.7783375314861463,133 "grad_norm": 4.468409061431885,134 "learning_rate": 9e-05,135 "loss": 7.2643,136 "step": 9000137 },138 {139 "epoch": 3.988245172124265,140 "grad_norm": 4.664194107055664,141 "learning_rate": 9.5e-05,142 "loss": 7.2599,143 "step": 9500144 },145 {146 "epoch": 4.198152812762385,147 "grad_norm": 4.949222564697266,148 "learning_rate": 0.0001,149 "loss": 7.0906,150 "step": 10000151 },152 {153 "epoch": 4.198152812762385,154 "eval_accuracy": 0.14933637552519546,155 "eval_loss": 7.23159122467041,156 "eval_runtime": 1.0708,157 "eval_samples_per_second": 447.345,158 "eval_steps_per_second": 4.67,159 "step": 10000160 },161 {162 "epoch": 4.408060453400504,163 "grad_norm": 4.300989627838135,164 "learning_rate": 9.899091826437942e-05,165 "loss": 7.1063,166 "step": 10500167 },168 {169 "epoch": 4.617968094038623,170 "grad_norm": 4.733574390411377,171 "learning_rate": 9.798183652875883e-05,172 "loss": 7.1007,173 "step": 11000174 },175 {176 "epoch": 4.827875734676742,177 "grad_norm": 5.090231418609619,178 "learning_rate": 9.697275479313825e-05,179 "loss": 7.0946,180 "step": 11500181 },182 {183 "epoch": 5.037783375314861,184 "grad_norm": 3.6812405586242676,185 "learning_rate": 9.596367305751767e-05,186 "loss": 7.0535,187 "step": 12000188 },189 {190 "epoch": 5.247691015952981,191 "grad_norm": 3.7199642658233643,192 "learning_rate": 9.495459132189707e-05,193 "loss": 6.9155,194 "step": 12500195 },196 {197 "epoch": 5.4575986565911,198 "grad_norm": 3.8014750480651855,199 "learning_rate": 9.394550958627649e-05,200 "loss": 6.9244,201 "step": 13000202 },203 {204 "epoch": 5.667506297229219,205 "grad_norm": 4.38668155670166,206 "learning_rate": 9.293642785065591e-05,207 "loss": 6.9336,208 "step": 13500209 },210 {211 "epoch": 5.877413937867338,212 "grad_norm": 4.114847183227539,213 "learning_rate": 9.192734611503532e-05,214 "loss": 6.9379,215 "step": 14000216 },217 {218 "epoch": 6.087321578505458,219 "grad_norm": 3.991894006729126,220 "learning_rate": 9.091826437941474e-05,221 "loss": 6.8481,222 "step": 14500223 },224 {225 "epoch": 6.297229219143577,226 "grad_norm": 3.796557903289795,227 "learning_rate": 8.990918264379415e-05,228 "loss": 6.7738,229 "step": 15000230 },231 {232 "epoch": 6.507136859781696,233 "grad_norm": 4.997390270233154,234 "learning_rate": 8.890010090817356e-05,235 "loss": 6.7876,236 "step": 15500237 },238 {239 "epoch": 6.717044500419815,240 "grad_norm": 3.105367422103882,241 "learning_rate": 8.789101917255298e-05,242 "loss": 6.8055,243 "step": 16000244 },245 {246 "epoch": 6.926952141057934,247 "grad_norm": 3.6604092121124268,248 "learning_rate": 8.68819374369324e-05,249 "loss": 6.811,250 "step": 16500251 },252 {253 "epoch": 7.136859781696054,254 "grad_norm": 4.096405029296875,255 "learning_rate": 8.58728557013118e-05,256 "loss": 6.6924,257 "step": 17000258 },259 {260 "epoch": 7.346767422334173,261 "grad_norm": 3.752100706100464,262 "learning_rate": 8.486377396569122e-05,263 "loss": 6.6527,264 "step": 17500265 },266 {267 "epoch": 7.5566750629722925,268 "grad_norm": 3.462996244430542,269 "learning_rate": 8.385469223007064e-05,270 "loss": 6.6678,271 "step": 18000272 },273 {274 "epoch": 7.766582703610411,275 "grad_norm": 3.705519437789917,276 "learning_rate": 8.284561049445006e-05,277 "loss": 6.6887,278 "step": 18500279 },280 {281 "epoch": 7.97649034424853,282 "grad_norm": 4.445788860321045,283 "learning_rate": 8.183652875882947e-05,284 "loss": 6.6944,285 "step": 19000286 },287 {288 "epoch": 8.18639798488665,289 "grad_norm": 3.868389368057251,290 "learning_rate": 8.082744702320889e-05,291 "loss": 6.5522,292 "step": 19500293 },294 {295 "epoch": 8.39630562552477,296 "grad_norm": 3.729473829269409,297 "learning_rate": 7.98183652875883e-05,298 "loss": 6.5625,299 "step": 20000300 },301 {302 "epoch": 8.39630562552477,303 "eval_accuracy": 0.1526137097000477,304 "eval_loss": 7.036741733551025,305 "eval_runtime": 1.0589,306 "eval_samples_per_second": 452.355,307 "eval_steps_per_second": 4.722,308 "step": 20000309 },310 {311 "epoch": 8.606213266162888,312 "grad_norm": 3.6588330268859863,313 "learning_rate": 7.880928355196771e-05,314 "loss": 6.5614,315 "step": 20500316 },317 {318 "epoch": 8.816120906801007,319 "grad_norm": 4.638293743133545,320 "learning_rate": 7.780020181634713e-05,321 "loss": 6.5764,322 "step": 21000323 },324 {325 "epoch": 9.026028547439127,326 "grad_norm": 4.317777633666992,327 "learning_rate": 7.679112008072655e-05,328 "loss": 6.5686,329 "step": 21500330 },331 {332 "epoch": 9.235936188077247,333 "grad_norm": 4.681490898132324,334 "learning_rate": 7.578203834510595e-05,335 "loss": 6.4478,336 "step": 22000337 },338 {339 "epoch": 9.445843828715365,340 "grad_norm": 4.315625190734863,341 "learning_rate": 7.477295660948537e-05,342 "loss": 6.4594,343 "step": 22500344 },345 {346 "epoch": 9.655751469353484,347 "grad_norm": 3.847973346710205,348 "learning_rate": 7.376387487386479e-05,349 "loss": 6.4783,350 "step": 23000351 },352 {353 "epoch": 9.865659109991604,354 "grad_norm": 6.255940914154053,355 "learning_rate": 7.27547931382442e-05,356 "loss": 6.4777,357 "step": 23500358 },359 {360 "epoch": 10.075566750629722,361 "grad_norm": 3.730644941329956,362 "learning_rate": 7.174571140262362e-05,363 "loss": 6.4379,364 "step": 24000365 },366 {367 "epoch": 10.285474391267842,368 "grad_norm": 3.6163721084594727,369 "learning_rate": 7.073662966700304e-05,370 "loss": 6.3757,371 "step": 24500372 },373 {374 "epoch": 10.495382031905962,375 "grad_norm": 6.299481391906738,376 "learning_rate": 6.972754793138244e-05,377 "loss": 6.3724,378 "step": 25000379 },380 {381 "epoch": 10.705289672544081,382 "grad_norm": 4.287810802459717,383 "learning_rate": 6.871846619576186e-05,384 "loss": 6.3921,385 "step": 25500386 },387 {388 "epoch": 10.9151973131822,389 "grad_norm": 4.100649356842041,390 "learning_rate": 6.770938446014128e-05,391 "loss": 6.402,392 "step": 26000393 },394 {395 "epoch": 11.125104953820319,396 "grad_norm": 4.9406819343566895,397 "learning_rate": 6.670030272452068e-05,398 "loss": 6.3409,399 "step": 26500400 },401 {402 "epoch": 11.335012594458439,403 "grad_norm": 3.7887113094329834,404 "learning_rate": 6.56912209889001e-05,405 "loss": 6.3122,406 "step": 27000407 },408 {409 "epoch": 11.544920235096557,410 "grad_norm": 2.6583824157714844,411 "learning_rate": 6.468213925327952e-05,412 "loss": 6.3141,413 "step": 27500414 },415 {416 "epoch": 11.754827875734676,417 "grad_norm": 4.447488784790039,418 "learning_rate": 6.367305751765893e-05,419 "loss": 6.3233,420 "step": 28000421 },422 {423 "epoch": 11.964735516372796,424 "grad_norm": 3.666630983352661,425 "learning_rate": 6.266397578203835e-05,426 "loss": 6.3288,427 "step": 28500428 },429 {430 "epoch": 12.174643157010916,431 "grad_norm": 5.452030658721924,432 "learning_rate": 6.165489404641777e-05,433 "loss": 6.2633,434 "step": 29000435 },436 {437 "epoch": 12.384550797649034,438 "grad_norm": 4.218036651611328,439 "learning_rate": 6.064581231079718e-05,440 "loss": 6.2574,441 "step": 29500442 },443 {444 "epoch": 12.594458438287154,445 "grad_norm": 2.8640193939208984,446 "learning_rate": 5.963673057517659e-05,447 "loss": 6.2634,448 "step": 30000449 },450 {451 "epoch": 12.594458438287154,452 "eval_accuracy": 0.15451772672580238,453 "eval_loss": 7.229600429534912,454 "eval_runtime": 1.0176,455 "eval_samples_per_second": 470.714,456 "eval_steps_per_second": 4.914,457 "step": 30000458 },459 {460 "epoch": 12.804366078925273,461 "grad_norm": 5.212186813354492,462 "learning_rate": 5.8627648839556e-05,463 "loss": 6.2586,464 "step": 30500465 },466 {467 "epoch": 13.014273719563391,468 "grad_norm": 3.6361076831817627,469 "learning_rate": 5.7618567103935416e-05,470 "loss": 6.2629,471 "step": 31000472 },473 {474 "epoch": 13.224181360201511,475 "grad_norm": 5.413948059082031,476 "learning_rate": 5.660948536831484e-05,477 "loss": 6.206,478 "step": 31500479 },480 {481 "epoch": 13.43408900083963,482 "grad_norm": 4.079280853271484,483 "learning_rate": 5.5600403632694254e-05,484 "loss": 6.213,485 "step": 32000486 },487 {488 "epoch": 13.64399664147775,489 "grad_norm": 4.215555667877197,490 "learning_rate": 5.4591321897073666e-05,491 "loss": 6.2147,492 "step": 32500493 },494 {495 "epoch": 13.853904282115868,496 "grad_norm": 3.849573850631714,497 "learning_rate": 5.358224016145308e-05,498 "loss": 6.2189,499 "step": 33000500 },501 {502 "epoch": 14.063811922753988,503 "grad_norm": 4.60481071472168,504 "learning_rate": 5.257315842583249e-05,505 "loss": 6.1962,506 "step": 33500507 },508 {509 "epoch": 14.273719563392108,510 "grad_norm": 6.386091232299805,511 "learning_rate": 5.15640766902119e-05,512 "loss": 6.1677,513 "step": 34000514 },515 {516 "epoch": 14.483627204030226,517 "grad_norm": 4.499414443969727,518 "learning_rate": 5.055499495459133e-05,519 "loss": 6.1705,520 "step": 34500521 },522 {523 "epoch": 14.693534844668346,524 "grad_norm": 3.7441251277923584,525 "learning_rate": 4.954591321897074e-05,526 "loss": 6.1722,527 "step": 35000528 },529 {530 "epoch": 14.903442485306465,531 "grad_norm": 3.144296646118164,532 "learning_rate": 4.853683148335015e-05,533 "loss": 6.1723,534 "step": 35500535 },536 {537 "epoch": 15.113350125944585,538 "grad_norm": 3.7621850967407227,539 "learning_rate": 4.7527749747729566e-05,540 "loss": 6.1493,541 "step": 36000542 },543 {544 "epoch": 15.323257766582703,545 "grad_norm": 3.3722171783447266,546 "learning_rate": 4.6518668012108985e-05,547 "loss": 6.1382,548 "step": 36500549 },550 {551 "epoch": 15.533165407220823,552 "grad_norm": 4.0898237228393555,553 "learning_rate": 4.55095862764884e-05,554 "loss": 6.1344,555 "step": 37000556 },557 {558 "epoch": 15.743073047858942,559 "grad_norm": 2.528909206390381,560 "learning_rate": 4.450050454086781e-05,561 "loss": 6.1404,562 "step": 37500563 },564 {565 "epoch": 15.95298068849706,566 "grad_norm": 3.6384353637695312,567 "learning_rate": 4.349142280524723e-05,568 "loss": 6.1333,569 "step": 38000570 },571 {572 "epoch": 16.162888329135182,573 "grad_norm": 4.224349021911621,574 "learning_rate": 4.248234106962664e-05,575 "loss": 6.1154,576 "step": 38500577 },578 {579 "epoch": 16.3727959697733,580 "grad_norm": 2.7164251804351807,581 "learning_rate": 4.147325933400605e-05,582 "loss": 6.1076,583 "step": 39000584 },585 {586 "epoch": 16.582703610411418,587 "grad_norm": 3.8217499256134033,588 "learning_rate": 4.046417759838547e-05,589 "loss": 6.109,590 "step": 39500591 },592 {593 "epoch": 16.79261125104954,594 "grad_norm": 4.337832927703857,595 "learning_rate": 3.9455095862764884e-05,596 "loss": 6.0994,597 "step": 40000598 },599 {600 "epoch": 16.79261125104954,601 "eval_accuracy": 0.15229337221038466,602 "eval_loss": 7.491790294647217,603 "eval_runtime": 1.0187,604 "eval_samples_per_second": 470.228,605 "eval_steps_per_second": 4.908,606 "step": 40000607 },608 {609 "epoch": 17.002518891687657,610 "grad_norm": 3.712265729904175,611 "learning_rate": 3.84460141271443e-05,612 "loss": 6.1037,613 "step": 40500614 },615 {616 "epoch": 17.212426532325775,617 "grad_norm": 3.09163761138916,618 "learning_rate": 3.7436932391523715e-05,619 "loss": 6.0796,620 "step": 41000621 },622 {623 "epoch": 17.422334172963897,624 "grad_norm": 3.670301675796509,625 "learning_rate": 3.642785065590313e-05,626 "loss": 6.0753,627 "step": 41500628 },629 {630 "epoch": 17.632241813602015,631 "grad_norm": 3.4407198429107666,632 "learning_rate": 3.541876892028255e-05,633 "loss": 6.0812,634 "step": 42000635 },636 {637 "epoch": 17.842149454240133,638 "grad_norm": 2.8078510761260986,639 "learning_rate": 3.440968718466196e-05,640 "loss": 6.0788,641 "step": 42500642 },643 {644 "epoch": 18.052057094878254,645 "grad_norm": 2.582552194595337,646 "learning_rate": 3.340060544904137e-05,647 "loss": 6.0706,648 "step": 43000649 },650 {651 "epoch": 18.261964735516372,652 "grad_norm": 3.352553606033325,653 "learning_rate": 3.239152371342079e-05,654 "loss": 6.0535,655 "step": 43500656 },657 {658 "epoch": 18.471872376154494,659 "grad_norm": 1.4775241613388062,660 "learning_rate": 3.13824419778002e-05,661 "loss": 6.0499,662 "step": 44000663 },664 {665 "epoch": 18.68178001679261,666 "grad_norm": 2.1691818237304688,667 "learning_rate": 3.0373360242179615e-05,668 "loss": 6.0492,669 "step": 44500670 },671 {672 "epoch": 18.89168765743073,673 "grad_norm": 3.906991720199585,674 "learning_rate": 2.9364278506559034e-05,675 "loss": 6.0534,676 "step": 45000677 },678 {679 "epoch": 19.10159529806885,680 "grad_norm": 4.238534927368164,681 "learning_rate": 2.8355196770938446e-05,682 "loss": 6.0443,683 "step": 45500684 },685 {686 "epoch": 19.31150293870697,687 "grad_norm": 3.73866868019104,688 "learning_rate": 2.734611503531786e-05,689 "loss": 6.0267,690 "step": 46000691 },692 {693 "epoch": 19.521410579345087,694 "grad_norm": 4.061879634857178,695 "learning_rate": 2.6337033299697278e-05,696 "loss": 6.029,697 "step": 46500698 },699 {700 "epoch": 19.73131821998321,701 "grad_norm": 1.6275923252105713,702 "learning_rate": 2.532795156407669e-05,703 "loss": 6.0312,704 "step": 47000705 },706 {707 "epoch": 19.941225860621326,708 "grad_norm": 2.3194937705993652,709 "learning_rate": 2.4318869828456106e-05,710 "loss": 6.024,711 "step": 47500712 },713 {714 "epoch": 20.151133501259444,715 "grad_norm": 2.4898221492767334,716 "learning_rate": 2.330978809283552e-05,717 "loss": 6.0184,718 "step": 48000719 },720 {721 "epoch": 20.361041141897566,722 "grad_norm": 1.5689467191696167,723 "learning_rate": 2.2300706357214934e-05,724 "loss": 6.0113,725 "step": 48500726 },727 {728 "epoch": 20.570948782535684,729 "grad_norm": 2.0924572944641113,730 "learning_rate": 2.129162462159435e-05,731 "loss": 6.0047,732 "step": 49000733 },734 {735 "epoch": 20.7808564231738,736 "grad_norm": 3.0981178283691406,737 "learning_rate": 2.0282542885973765e-05,738 "loss": 6.0064,739 "step": 49500740 },741 {742 "epoch": 20.990764063811923,743 "grad_norm": 4.074100494384766,744 "learning_rate": 1.927346115035318e-05,745 "loss": 6.0062,746 "step": 50000747 },748 {749 "epoch": 20.990764063811923,750 "eval_accuracy": 0.1510920458687302,751 "eval_loss": 7.390818119049072,752 "eval_runtime": 1.0237,753 "eval_samples_per_second": 467.917,754 "eval_steps_per_second": 4.884,755 "step": 50000756 },757 {758 "epoch": 21.20067170445004,759 "grad_norm": 2.681124687194824,760 "learning_rate": 1.8264379414732593e-05,761 "loss": 5.9937,762 "step": 50500763 },764 {765 "epoch": 21.410579345088163,766 "grad_norm": 2.0725343227386475,767 "learning_rate": 1.725529767911201e-05,768 "loss": 5.9944,769 "step": 51000770 },771 {772 "epoch": 21.62048698572628,773 "grad_norm": 5.385054111480713,774 "learning_rate": 1.6246215943491424e-05,775 "loss": 5.9913,776 "step": 51500777 },778 {779 "epoch": 21.8303946263644,780 "grad_norm": 1.695888876914978,781 "learning_rate": 1.5237134207870837e-05,782 "loss": 5.9879,783 "step": 52000784 },785 {786 "epoch": 22.04030226700252,787 "grad_norm": 2.4581727981567383,788 "learning_rate": 1.4228052472250252e-05,789 "loss": 5.9876,790 "step": 52500791 },792 {793 "epoch": 22.250209907640638,794 "grad_norm": 0.8789335489273071,795 "learning_rate": 1.3218970736629668e-05,796 "loss": 5.9758,797 "step": 53000798 },799 {800 "epoch": 22.460117548278756,801 "grad_norm": 6.385268211364746,802 "learning_rate": 1.2209889001009082e-05,803 "loss": 5.9773,804 "step": 53500805 },806 {807 "epoch": 22.670025188916878,808 "grad_norm": 3.9702112674713135,809 "learning_rate": 1.1200807265388496e-05,810 "loss": 5.9772,811 "step": 54000812 },813 {814 "epoch": 22.879932829554996,815 "grad_norm": 3.841106653213501,816 "learning_rate": 1.0191725529767912e-05,817 "loss": 5.9728,818 "step": 54500819 },820 {821 "epoch": 23.089840470193113,822 "grad_norm": 2.1624410152435303,823 "learning_rate": 9.182643794147326e-06,824 "loss": 5.9677,825 "step": 55000826 },827 {828 "epoch": 23.299748110831235,829 "grad_norm": 1.8635454177856445,830 "learning_rate": 8.173562058526741e-06,831 "loss": 5.9647,832 "step": 55500833 },834 {835 "epoch": 23.509655751469353,836 "grad_norm": 1.5375641584396362,837 "learning_rate": 7.164480322906156e-06,838 "loss": 5.9643,839 "step": 56000840 },841 {842 "epoch": 23.719563392107474,843 "grad_norm": 2.0004587173461914,844 "learning_rate": 6.155398587285571e-06,845 "loss": 5.9589,846 "step": 56500847 },848 {849 "epoch": 23.929471032745592,850 "grad_norm": 4.578677654266357,851 "learning_rate": 5.146316851664986e-06,852 "loss": 5.9608,853 "step": 57000854 },855 {856 "epoch": 24.13937867338371,857 "grad_norm": 2.199143171310425,858 "learning_rate": 4.1372351160444e-06,859 "loss": 5.9534,860 "step": 57500861 },862 {863 "epoch": 24.349286314021832,864 "grad_norm": 0.8375799655914307,865 "learning_rate": 3.1281533804238145e-06,866 "loss": 5.9541,867 "step": 58000868 },869 {870 "epoch": 24.55919395465995,871 "grad_norm": 3.7426412105560303,872 "learning_rate": 2.1190716448032293e-06,873 "loss": 5.9519,874 "step": 58500875 },876 {877 "epoch": 24.769101595298068,878 "grad_norm": 0.7772536277770996,879 "learning_rate": 1.109989909182644e-06,880 "loss": 5.9495,881 "step": 59000882 },883 {884 "epoch": 24.97900923593619,885 "grad_norm": 0.813712477684021,886 "learning_rate": 1.0090817356205853e-07,887 "loss": 5.9498,888 "step": 59500889 },890 {891 "epoch": 25.0,892 "step": 59550,893 "total_flos": 1.5161929390646784e+18,894 "train_loss": 6.650803347186418,895 "train_runtime": 24926.7454,896 "train_samples_per_second": 229.311,897 "train_steps_per_second": 2.389898 }899 ],900 "logging_steps": 500,901 "max_steps": 59550,902 "num_input_tokens_seen": 0,903 "num_train_epochs": 25,904 "save_steps": 10000,905 "stateful_callbacks": {906 "TrainerControl": {907 "args": {908 "should_epoch_stop": false,909 "should_evaluate": false,910 "should_log": false,911 "should_save": true,912 "should_training_stop": true913 },914 "attributes": {}915 }916 },917 "total_flos": 1.5161929390646784e+18,918 "train_batch_size": 96,919 "trial_name": null,920 "trial_params": null921}922 