EleutherAI/less-replication-7b-warmup
110
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 4.0,6 "eval_steps": 500,7 "global_step": 424,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "entropy": 1.7230634093284607,14 "epoch": 0.009478672985781991,15 "grad_norm": 1.5876940488815308,16 "learning_rate": 0.0,17 "loss": 1.6636556386947632,18 "mean_token_accuracy": 0.6497756391763687,19 "num_tokens": 16375.0,20 "step": 121 },22 {23 "entropy": 1.7448452711105347,24 "epoch": 0.018957345971563982,25 "grad_norm": 1.609075665473938,26 "learning_rate": 9.090909090909091e-07,27 "loss": 1.7037649154663086,28 "mean_token_accuracy": 0.6527450829744339,29 "num_tokens": 31738.0,30 "step": 231 },32 {33 "entropy": 1.710326075553894,34 "epoch": 0.02843601895734597,35 "grad_norm": 1.7256412506103516,36 "learning_rate": 1.8181818181818183e-06,37 "loss": 1.663683533668518,38 "mean_token_accuracy": 0.6594014763832092,39 "num_tokens": 48323.0,40 "step": 341 },42 {43 "entropy": 1.8404302299022675,44 "epoch": 0.037914691943127965,45 "grad_norm": 1.7094064950942993,46 "learning_rate": 2.7272727272727272e-06,47 "loss": 1.8555431365966797,48 "mean_token_accuracy": 0.6288548111915588,49 "num_tokens": 63474.0,50 "step": 451 },52 {53 "entropy": 1.61773881316185,54 "epoch": 0.04739336492890995,55 "grad_norm": 1.4093204736709595,56 "learning_rate": 3.6363636363636366e-06,57 "loss": 1.586500883102417,58 "mean_token_accuracy": 0.6745017617940903,59 "num_tokens": 81340.0,60 "step": 561 },62 {63 "entropy": 1.713468849658966,64 "epoch": 0.05687203791469194,65 "grad_norm": 1.4611302614212036,66 "learning_rate": 4.5454545454545455e-06,67 "loss": 1.693520188331604,68 "mean_token_accuracy": 0.6446037888526917,69 "num_tokens": 98224.0,70 "step": 671 },72 {73 "entropy": 1.7964849770069122,74 "epoch": 0.06635071090047394,75 "grad_norm": 2.0188069343566895,76 "learning_rate": 5.4545454545454545e-06,77 "loss": 1.7727513313293457,78 "mean_token_accuracy": 0.637175589799881,79 "num_tokens": 113456.0,80 "step": 781 },82 {83 "entropy": 1.8277963697910309,84 "epoch": 0.07582938388625593,85 "grad_norm": 2.102283000946045,86 "learning_rate": 6.363636363636364e-06,87 "loss": 1.74727463722229,88 "mean_token_accuracy": 0.6403038948774338,89 "num_tokens": 129166.0,90 "step": 891 },92 {93 "entropy": 1.7185819745063782,94 "epoch": 0.08530805687203792,95 "grad_norm": 1.5833488702774048,96 "learning_rate": 7.272727272727273e-06,97 "loss": 1.653383731842041,98 "mean_token_accuracy": 0.6527138650417328,99 "num_tokens": 145103.0,100 "step": 9101 },102 {103 "entropy": 1.7715328633785248,104 "epoch": 0.0947867298578199,105 "grad_norm": 1.4368863105773926,106 "learning_rate": 8.181818181818183e-06,107 "loss": 1.6846437454223633,108 "mean_token_accuracy": 0.6343469023704529,109 "num_tokens": 160889.0,110 "step": 10111 },112 {113 "entropy": 1.6594811379909515,114 "epoch": 0.10426540284360189,115 "grad_norm": 1.1473579406738281,116 "learning_rate": 9.090909090909091e-06,117 "loss": 1.5402858257293701,118 "mean_token_accuracy": 0.663172259926796,119 "num_tokens": 177508.0,120 "step": 11121 },122 {123 "entropy": 1.6768087446689606,124 "epoch": 0.11374407582938388,125 "grad_norm": 1.0232007503509521,126 "learning_rate": 1e-05,127 "loss": 1.5721584558486938,128 "mean_token_accuracy": 0.6572738438844681,129 "num_tokens": 194325.0,130 "step": 12131 },132 {133 "entropy": 1.7810584902763367,134 "epoch": 0.12322274881516587,135 "grad_norm": 1.0588256120681763,136 "learning_rate": 1.0909090909090909e-05,137 "loss": 1.6563284397125244,138 "mean_token_accuracy": 0.6360882222652435,139 "num_tokens": 209935.0,140 "step": 13141 },142 {143 "entropy": 1.6035055816173553,144 "epoch": 0.13270142180094788,145 "grad_norm": 0.9055944681167603,146 "learning_rate": 1.181818181818182e-05,147 "loss": 1.4942286014556885,148 "mean_token_accuracy": 0.6743725687265396,149 "num_tokens": 226137.0,150 "step": 14151 },152 {153 "entropy": 1.5955874025821686,154 "epoch": 0.14218009478672985,155 "grad_norm": 0.8588143587112427,156 "learning_rate": 1.2727272727272728e-05,157 "loss": 1.4966249465942383,158 "mean_token_accuracy": 0.66853067278862,159 "num_tokens": 241952.0,160 "step": 15161 },162 {163 "entropy": 1.607325166463852,164 "epoch": 0.15165876777251186,165 "grad_norm": 0.9346417784690857,166 "learning_rate": 1.3636363636363637e-05,167 "loss": 1.4750392436981201,168 "mean_token_accuracy": 0.660357192158699,169 "num_tokens": 257481.0,170 "step": 16171 },172 {173 "entropy": 1.5410204827785492,174 "epoch": 0.16113744075829384,175 "grad_norm": 0.8730961680412292,176 "learning_rate": 1.4545454545454546e-05,177 "loss": 1.4322866201400757,178 "mean_token_accuracy": 0.6814255267381668,179 "num_tokens": 273287.0,180 "step": 17181 },182 {183 "entropy": 1.5345198810100555,184 "epoch": 0.17061611374407584,185 "grad_norm": 0.8538835048675537,186 "learning_rate": 1.5454545454545454e-05,187 "loss": 1.3966718912124634,188 "mean_token_accuracy": 0.6867176294326782,189 "num_tokens": 288809.0,190 "step": 18191 },192 {193 "entropy": 1.4507653713226318,194 "epoch": 0.18009478672985782,195 "grad_norm": 0.7879626154899597,196 "learning_rate": 1.6363636363636366e-05,197 "loss": 1.3630776405334473,198 "mean_token_accuracy": 0.6779211610555649,199 "num_tokens": 304801.0,200 "step": 19201 },202 {203 "entropy": 1.4710645079612732,204 "epoch": 0.1895734597156398,205 "grad_norm": 0.8410776257514954,206 "learning_rate": 1.7272727272727274e-05,207 "loss": 1.4051768779754639,208 "mean_token_accuracy": 0.6760597079992294,209 "num_tokens": 320246.0,210 "step": 20211 },212 {213 "entropy": 1.3435666263103485,214 "epoch": 0.1990521327014218,215 "grad_norm": 0.759080708026886,216 "learning_rate": 1.8181818181818182e-05,217 "loss": 1.267665147781372,218 "mean_token_accuracy": 0.7003267407417297,219 "num_tokens": 337053.0,220 "step": 21221 },222 {223 "entropy": 1.394632339477539,224 "epoch": 0.20853080568720378,225 "grad_norm": 0.8370321989059448,226 "learning_rate": 1.9090909090909094e-05,227 "loss": 1.3486745357513428,228 "mean_token_accuracy": 0.6893624067306519,229 "num_tokens": 352268.0,230 "step": 22231 },232 {233 "entropy": 1.1503786146640778,234 "epoch": 0.21800947867298578,235 "grad_norm": 0.7440245747566223,236 "learning_rate": 2e-05,237 "loss": 1.1109555959701538,238 "mean_token_accuracy": 0.7320543378591537,239 "num_tokens": 370034.0,240 "step": 23241 },242 {243 "entropy": 1.1959196627140045,244 "epoch": 0.22748815165876776,245 "grad_norm": 0.8951184749603271,246 "learning_rate": 1.9999694637689328e-05,247 "loss": 1.170793056488037,248 "mean_token_accuracy": 0.7196634411811829,249 "num_tokens": 385616.0,250 "step": 24251 },252 {253 "entropy": 1.1576418280601501,254 "epoch": 0.23696682464454977,255 "grad_norm": 0.8570479154586792,256 "learning_rate": 1.999877856940653e-05,257 "loss": 1.1886144876480103,258 "mean_token_accuracy": 0.7251218110322952,259 "num_tokens": 401820.0,260 "step": 25261 },262 {263 "entropy": 1.1380924582481384,264 "epoch": 0.24644549763033174,265 "grad_norm": 0.8543166518211365,266 "learning_rate": 1.999725185109816e-05,267 "loss": 1.2082655429840088,268 "mean_token_accuracy": 0.711901068687439,269 "num_tokens": 417425.0,270 "step": 26271 },272 {273 "entropy": 0.9601393789052963,274 "epoch": 0.2559241706161137,275 "grad_norm": 0.863011360168457,276 "learning_rate": 1.999511457600466e-05,277 "loss": 1.0240533351898193,278 "mean_token_accuracy": 0.7590129524469376,279 "num_tokens": 433582.0,280 "step": 27281 },282 {283 "entropy": 1.1091893315315247,284 "epoch": 0.26540284360189575,285 "grad_norm": 0.8494651317596436,286 "learning_rate": 1.9992366874654684e-05,287 "loss": 1.2075942754745483,288 "mean_token_accuracy": 0.7092257142066956,289 "num_tokens": 448489.0,290 "step": 28291 },292 {293 "entropy": 1.0761875212192535,294 "epoch": 0.27488151658767773,295 "grad_norm": 0.7775688767433167,296 "learning_rate": 1.9989008914857115e-05,297 "loss": 1.1314418315887451,298 "mean_token_accuracy": 0.7183841317892075,299 "num_tokens": 465063.0,300 "step": 29301 },302 {303 "entropy": 1.0514411181211472,304 "epoch": 0.2843601895734597,305 "grad_norm": 0.7758051753044128,306 "learning_rate": 1.998504090169083e-05,307 "loss": 1.1378644704818726,308 "mean_token_accuracy": 0.7295740246772766,309 "num_tokens": 481645.0,310 "step": 30311 },312 {313 "entropy": 1.0378742814064026,314 "epoch": 0.2938388625592417,315 "grad_norm": 0.7737782597541809,316 "learning_rate": 1.998046307749216e-05,317 "loss": 1.0405387878417969,318 "mean_token_accuracy": 0.7435884028673172,319 "num_tokens": 498020.0,320 "step": 31321 },322 {323 "entropy": 1.162422239780426,324 "epoch": 0.3033175355450237,325 "grad_norm": 0.8285279870033264,326 "learning_rate": 1.9975275721840105e-05,327 "loss": 1.2049440145492554,328 "mean_token_accuracy": 0.7072447687387466,329 "num_tokens": 514133.0,330 "step": 32331 },332 {333 "entropy": 1.0860552787780762,334 "epoch": 0.3127962085308057,335 "grad_norm": 0.7870988249778748,336 "learning_rate": 1.9969479151539238e-05,337 "loss": 1.088366985321045,338 "mean_token_accuracy": 0.7330214083194733,339 "num_tokens": 529902.0,340 "step": 33341 },342 {343 "entropy": 1.0763448625802994,344 "epoch": 0.3222748815165877,345 "grad_norm": 0.8604663610458374,346 "learning_rate": 1.9963073720600383e-05,347 "loss": 1.0562834739685059,348 "mean_token_accuracy": 0.7405408322811127,349 "num_tokens": 545365.0,350 "step": 34351 },352 {353 "entropy": 1.0641998052597046,354 "epoch": 0.33175355450236965,355 "grad_norm": 0.8490760922431946,356 "learning_rate": 1.9956059820218982e-05,357 "loss": 1.08937668800354,358 "mean_token_accuracy": 0.7371959090232849,359 "num_tokens": 560629.0,360 "step": 35361 },362 {363 "entropy": 1.024262085556984,364 "epoch": 0.3412322274881517,365 "grad_norm": 0.7459779381752014,366 "learning_rate": 1.99484378787512e-05,367 "loss": 1.0313022136688232,368 "mean_token_accuracy": 0.7482277005910873,369 "num_tokens": 575979.0,370 "step": 36371 },372 {373 "entropy": 1.0767400860786438,374 "epoch": 0.35071090047393366,375 "grad_norm": 0.6798542141914368,376 "learning_rate": 1.9940208361687762e-05,377 "loss": 1.1029484272003174,378 "mean_token_accuracy": 0.7400920391082764,379 "num_tokens": 593219.0,380 "step": 37381 },382 {383 "entropy": 1.0605872124433517,384 "epoch": 0.36018957345971564,385 "grad_norm": 0.775074303150177,386 "learning_rate": 1.9931371771625545e-05,387 "loss": 1.1065254211425781,388 "mean_token_accuracy": 0.7417906671762466,389 "num_tokens": 608173.0,390 "step": 38391 },392 {393 "entropy": 1.002692699432373,394 "epoch": 0.3696682464454976,395 "grad_norm": 0.7341852784156799,396 "learning_rate": 1.9921928648236855e-05,397 "loss": 1.013192892074585,398 "mean_token_accuracy": 0.7493702918291092,399 "num_tokens": 624455.0,400 "step": 39401 },402 {403 "entropy": 1.0264743566513062,404 "epoch": 0.3791469194312796,405 "grad_norm": 0.709384024143219,406 "learning_rate": 1.9911879568236492e-05,407 "loss": 1.050273060798645,408 "mean_token_accuracy": 0.7467086464166641,409 "num_tokens": 639997.0,410 "step": 40411 },412 {413 "entropy": 0.9731618463993073,414 "epoch": 0.3886255924170616,415 "grad_norm": 0.7397781014442444,416 "learning_rate": 1.990122514534651e-05,417 "loss": 0.9792512655258179,418 "mean_token_accuracy": 0.7615833282470703,419 "num_tokens": 655244.0,420 "step": 41421 },422 {423 "entropy": 0.9393151998519897,424 "epoch": 0.3981042654028436,425 "grad_norm": 0.603070080280304,426 "learning_rate": 1.9889966030258752e-05,427 "loss": 0.9558578729629517,428 "mean_token_accuracy": 0.7628461569547653,429 "num_tokens": 672691.0,430 "step": 42431 },432 {433 "entropy": 0.9829478412866592,434 "epoch": 0.4075829383886256,435 "grad_norm": 0.736235499382019,436 "learning_rate": 1.9878102910595097e-05,437 "loss": 0.978938639163971,438 "mean_token_accuracy": 0.7605015784502029,439 "num_tokens": 688436.0,440 "step": 43441 },442 {443 "entropy": 0.9505813121795654,444 "epoch": 0.41706161137440756,445 "grad_norm": 0.6724395751953125,446 "learning_rate": 1.9865636510865466e-05,447 "loss": 0.978853702545166,448 "mean_token_accuracy": 0.762572169303894,449 "num_tokens": 705439.0,450 "step": 44451 },452 {453 "entropy": 0.9644436687231064,454 "epoch": 0.4265402843601896,455 "grad_norm": 0.7388243079185486,456 "learning_rate": 1.985256759242359e-05,457 "loss": 1.0140184164047241,458 "mean_token_accuracy": 0.7558823525905609,459 "num_tokens": 721153.0,460 "step": 45461 },462 {463 "entropy": 0.9982628226280212,464 "epoch": 0.43601895734597157,465 "grad_norm": 0.8184957504272461,466 "learning_rate": 1.9838896953420495e-05,467 "loss": 1.0390454530715942,468 "mean_token_accuracy": 0.7469227313995361,469 "num_tokens": 736548.0,470 "step": 46471 },472 {473 "entropy": 0.9748822003602982,474 "epoch": 0.44549763033175355,475 "grad_norm": 0.7347403764724731,476 "learning_rate": 1.982462542875576e-05,477 "loss": 0.9972015619277954,478 "mean_token_accuracy": 0.7542161792516708,479 "num_tokens": 753213.0,480 "step": 47481 },482 {483 "entropy": 0.9839194715023041,484 "epoch": 0.4549763033175355,485 "grad_norm": 0.7843495607376099,486 "learning_rate": 1.9809753890026543e-05,487 "loss": 0.9987256526947021,488 "mean_token_accuracy": 0.7523878663778305,489 "num_tokens": 768748.0,490 "step": 48491 },492 {493 "entropy": 1.0193238854408264,494 "epoch": 0.46445497630331756,495 "grad_norm": 0.7180641889572144,496 "learning_rate": 1.979428324547432e-05,497 "loss": 1.0118625164031982,498 "mean_token_accuracy": 0.7445009052753448,499 "num_tokens": 783698.0,500 "step": 49501 },502 {503 "entropy": 0.9688207358121872,504 "epoch": 0.47393364928909953,505 "grad_norm": 0.7978101372718811,506 "learning_rate": 1.9778214439929453e-05,507 "loss": 0.9732460379600525,508 "mean_token_accuracy": 0.7674588412046432,509 "num_tokens": 799825.0,510 "step": 50511 },512 {513 "entropy": 1.000299409031868,514 "epoch": 0.4834123222748815,515 "grad_norm": 0.7578161358833313,516 "learning_rate": 1.9761548454753455e-05,517 "loss": 1.0048537254333496,518 "mean_token_accuracy": 0.7620798051357269,519 "num_tokens": 814971.0,520 "step": 51521 },522 {523 "entropy": 0.9922938197851181,524 "epoch": 0.4928909952606635,525 "grad_norm": 0.7843224406242371,526 "learning_rate": 1.9744286307779076e-05,527 "loss": 0.9955688714981079,528 "mean_token_accuracy": 0.7550618052482605,529 "num_tokens": 830524.0,530 "step": 52531 },532 {533 "entropy": 0.9540591835975647,534 "epoch": 0.5023696682464455,535 "grad_norm": 0.7341641783714294,536 "learning_rate": 1.972642905324813e-05,537 "loss": 0.9120223522186279,538 "mean_token_accuracy": 0.7745194137096405,539 "num_tokens": 845863.0,540 "step": 53541 },542 {543 "entropy": 0.9302634298801422,544 "epoch": 0.5118483412322274,545 "grad_norm": 0.7071745991706848,546 "learning_rate": 1.9707977781747126e-05,547 "loss": 0.8850457668304443,548 "mean_token_accuracy": 0.7787033319473267,549 "num_tokens": 861849.0,550 "step": 54551 },552 {553 "entropy": 0.9491895437240601,554 "epoch": 0.5213270142180095,555 "grad_norm": 0.826166033744812,556 "learning_rate": 1.9688933620140638e-05,557 "loss": 0.9278963804244995,558 "mean_token_accuracy": 0.7744901925325394,559 "num_tokens": 877161.0,560 "step": 55561 },562 {563 "entropy": 0.990895539522171,564 "epoch": 0.5308056872037915,565 "grad_norm": 0.94631028175354,566 "learning_rate": 1.966929773150251e-05,567 "loss": 0.9650328159332275,568 "mean_token_accuracy": 0.7605593204498291,569 "num_tokens": 892160.0,570 "step": 56571 },572 {573 "entropy": 0.9609559625387192,574 "epoch": 0.5402843601895735,575 "grad_norm": 0.6704272031784058,576 "learning_rate": 1.96490713150448e-05,577 "loss": 0.9601552486419678,578 "mean_token_accuracy": 0.7565764933824539,579 "num_tokens": 908720.0,580 "step": 57581 },582 {583 "entropy": 0.9724608361721039,584 "epoch": 0.5497630331753555,585 "grad_norm": 0.7987351417541504,586 "learning_rate": 1.9628255606044562e-05,587 "loss": 0.9856890439987183,588 "mean_token_accuracy": 0.7569890469312668,589 "num_tokens": 924429.0,590 "step": 58591 },592 {593 "entropy": 0.9434510767459869,594 "epoch": 0.5592417061611374,595 "grad_norm": 0.7899335622787476,596 "learning_rate": 1.9606851875768404e-05,597 "loss": 0.9372917413711548,598 "mean_token_accuracy": 0.7673724591732025,599 "num_tokens": 940816.0,600 "step": 59601 },602 {603 "entropy": 0.8971584439277649,604 "epoch": 0.5687203791469194,605 "grad_norm": 0.7808126211166382,606 "learning_rate": 1.9584861431394825e-05,607 "loss": 0.8875389099121094,608 "mean_token_accuracy": 0.7799597531557083,609 "num_tokens": 956724.0,610 "step": 60611 },612 {613 "entropy": 0.9393307864665985,614 "epoch": 0.5781990521327014,615 "grad_norm": 0.6976328492164612,616 "learning_rate": 1.956228561593441e-05,617 "loss": 0.9503510594367981,618 "mean_token_accuracy": 0.7596384882926941,619 "num_tokens": 973257.0,620 "step": 61621 },622 {623 "entropy": 1.021749958395958,624 "epoch": 0.5876777251184834,625 "grad_norm": 0.7356151342391968,626 "learning_rate": 1.953912580814779e-05,627 "loss": 1.0444157123565674,628 "mean_token_accuracy": 0.7519654631614685,629 "num_tokens": 990266.0,630 "step": 62631 },632 {633 "entropy": 0.8811668455600739,634 "epoch": 0.5971563981042654,635 "grad_norm": 0.8034309148788452,636 "learning_rate": 1.9515383422461457e-05,637 "loss": 0.888944685459137,638 "mean_token_accuracy": 0.77887162566185,639 "num_tokens": 1006065.0,640 "step": 63641 },642 {643 "entropy": 0.9695002734661102,644 "epoch": 0.6066350710900474,645 "grad_norm": 0.6834648251533508,646 "learning_rate": 1.949105990888135e-05,647 "loss": 0.9581116437911987,648 "mean_token_accuracy": 0.7590216845273972,649 "num_tokens": 1021785.0,650 "step": 64651 },652 {653 "entropy": 0.8993494361639023,654 "epoch": 0.6161137440758294,655 "grad_norm": 0.787246584892273,656 "learning_rate": 1.9466156752904344e-05,657 "loss": 0.874479353427887,658 "mean_token_accuracy": 0.7846818715333939,659 "num_tokens": 1037721.0,660 "step": 65661 },662 {663 "entropy": 0.9583495110273361,664 "epoch": 0.6255924170616114,665 "grad_norm": 0.7072925567626953,666 "learning_rate": 1.944067547542748e-05,667 "loss": 0.9677348136901855,668 "mean_token_accuracy": 0.7663833945989609,669 "num_tokens": 1054125.0,670 "step": 66671 },672 {673 "entropy": 0.926140233874321,674 "epoch": 0.6350710900473934,675 "grad_norm": 0.7638861536979675,676 "learning_rate": 1.9414617632655114e-05,677 "loss": 0.9518347382545471,678 "mean_token_accuracy": 0.7670052349567413,679 "num_tokens": 1070665.0,680 "step": 67681 },682 {683 "entropy": 0.9343065768480301,684 "epoch": 0.6445497630331753,685 "grad_norm": 0.8179787397384644,686 "learning_rate": 1.9387984816003868e-05,687 "loss": 0.9283621311187744,688 "mean_token_accuracy": 0.7648810148239136,689 "num_tokens": 1086022.0,690 "step": 68691 },692 {693 "entropy": 1.0115204006433487,694 "epoch": 0.6540284360189573,695 "grad_norm": 0.9361454844474792,696 "learning_rate": 1.9360778652005416e-05,697 "loss": 1.043932557106018,698 "mean_token_accuracy": 0.7457028925418854,699 "num_tokens": 1101344.0,700 "step": 69701 },702 {703 "entropy": 0.919638603925705,704 "epoch": 0.6635071090047393,705 "grad_norm": 0.7808894515037537,706 "learning_rate": 1.933300080220719e-05,707 "loss": 0.877983570098877,708 "mean_token_accuracy": 0.7854871153831482,709 "num_tokens": 1117702.0,710 "step": 70711 },712 {713 "entropy": 0.9225042313337326,714 "epoch": 0.6729857819905213,715 "grad_norm": 0.7920427918434143,716 "learning_rate": 1.9304652963070868e-05,717 "loss": 0.8995954990386963,718 "mean_token_accuracy": 0.7772288471460342,719 "num_tokens": 1134538.0,720 "step": 71721 },722 {723 "entropy": 0.9348872601985931,724 "epoch": 0.6824644549763034,725 "grad_norm": 0.7429097294807434,726 "learning_rate": 1.927573686586878e-05,727 "loss": 0.9216326475143433,728 "mean_token_accuracy": 0.7696940153837204,729 "num_tokens": 1150205.0,730 "step": 72731 },732 {733 "entropy": 0.8841167837381363,734 "epoch": 0.6919431279620853,735 "grad_norm": 0.9018954634666443,736 "learning_rate": 1.9246254276578175e-05,737 "loss": 0.8575382232666016,738 "mean_token_accuracy": 0.7889357209205627,739 "num_tokens": 1166758.0,740 "step": 73741 },742 {743 "entropy": 0.8816715627908707,744 "epoch": 0.7014218009478673,745 "grad_norm": 0.8096207976341248,746 "learning_rate": 1.9216206995773373e-05,747 "loss": 0.8846977949142456,748 "mean_token_accuracy": 0.7817238718271255,749 "num_tokens": 1182364.0,750 "step": 74751 },752 {753 "entropy": 0.9715549945831299,754 "epoch": 0.7109004739336493,755 "grad_norm": 0.9433501958847046,756 "learning_rate": 1.9185596858515797e-05,757 "loss": 0.9738811254501343,758 "mean_token_accuracy": 0.7634537816047668,759 "num_tokens": 1196839.0,760 "step": 75761 },762 {763 "entropy": 0.8470079302787781,764 "epoch": 0.7203791469194313,765 "grad_norm": 0.7724857926368713,766 "learning_rate": 1.9154425734241893e-05,767 "loss": 0.8701345920562744,768 "mean_token_accuracy": 0.78440722823143,769 "num_tokens": 1213377.0,770 "step": 76771 },772 {773 "entropy": 0.9654579311609268,774 "epoch": 0.7298578199052133,775 "grad_norm": 0.8752750158309937,776 "learning_rate": 1.9122695526648968e-05,777 "loss": 0.9896650910377502,778 "mean_token_accuracy": 0.752353623509407,779 "num_tokens": 1229571.0,780 "step": 77781 },782 {783 "entropy": 0.8873108923435211,784 "epoch": 0.7393364928909952,785 "grad_norm": 0.7646530866622925,786 "learning_rate": 1.9090408173578923e-05,787 "loss": 0.8805227279663086,788 "mean_token_accuracy": 0.7848268151283264,789 "num_tokens": 1247252.0,790 "step": 78791 },792 {793 "entropy": 0.8589294850826263,794 "epoch": 0.7488151658767772,795 "grad_norm": 0.8699224591255188,796 "learning_rate": 1.905756564689991e-05,797 "loss": 0.8414448499679565,798 "mean_token_accuracy": 0.7981384843587875,799 "num_tokens": 1264824.0,800 "step": 79801 },802 {803 "entropy": 0.9470502883195877,804 "epoch": 0.7582938388625592,805 "grad_norm": 0.7720149159431458,806 "learning_rate": 1.9024169952385887e-05,807 "loss": 0.9226930141448975,808 "mean_token_accuracy": 0.7769389748573303,809 "num_tokens": 1279205.0,810 "step": 80811 },812 {813 "entropy": 0.9329600483179092,814 "epoch": 0.7677725118483413,815 "grad_norm": 0.7501013875007629,816 "learning_rate": 1.8990223129594146e-05,817 "loss": 0.9248684644699097,818 "mean_token_accuracy": 0.7771976292133331,819 "num_tokens": 1295447.0,820 "step": 81821 },822 {823 "entropy": 0.9092026650905609,824 "epoch": 0.7772511848341233,825 "grad_norm": 0.7927464842796326,826 "learning_rate": 1.8955727251740742e-05,827 "loss": 0.9005250930786133,828 "mean_token_accuracy": 0.7794855386018753,829 "num_tokens": 1310786.0,830 "step": 82831 },832 {833 "entropy": 0.9609334319829941,834 "epoch": 0.7867298578199052,835 "grad_norm": 0.8200048804283142,836 "learning_rate": 1.8920684425573865e-05,837 "loss": 1.0015482902526855,838 "mean_token_accuracy": 0.7510274350643158,839 "num_tokens": 1326327.0,840 "step": 83841 },842 {843 "entropy": 0.9021405577659607,844 "epoch": 0.7962085308056872,845 "grad_norm": 0.7447288632392883,846 "learning_rate": 1.888509679124519e-05,847 "loss": 0.9190552234649658,848 "mean_token_accuracy": 0.7851650565862656,849 "num_tokens": 1342612.0,850 "step": 84851 },852 {853 "entropy": 0.8972304910421371,854 "epoch": 0.8056872037914692,855 "grad_norm": 0.8647413849830627,856 "learning_rate": 1.884896652217917e-05,857 "loss": 0.8718003034591675,858 "mean_token_accuracy": 0.7822871953248978,859 "num_tokens": 1358216.0,860 "step": 85861 },862 {863 "entropy": 0.8696665316820145,864 "epoch": 0.8151658767772512,865 "grad_norm": 0.8497427105903625,866 "learning_rate": 1.8812295824940284e-05,867 "loss": 0.8930065035820007,868 "mean_token_accuracy": 0.7775844633579254,869 "num_tokens": 1374136.0,870 "step": 86871 },872 {873 "entropy": 0.8540368229150772,874 "epoch": 0.8246445497630331,875 "grad_norm": 0.9805054068565369,876 "learning_rate": 1.877508693909831e-05,877 "loss": 0.8588962554931641,878 "mean_token_accuracy": 0.7839875668287277,879 "num_tokens": 1390236.0,880 "step": 87881 },882 {883 "entropy": 0.9288308620452881,884 "epoch": 0.8341232227488151,885 "grad_norm": 0.9268847107887268,886 "learning_rate": 1.8737342137091523e-05,887 "loss": 0.9457352161407471,888 "mean_token_accuracy": 0.7653509080410004,889 "num_tokens": 1405289.0,890 "step": 88891 },892 {893 "entropy": 0.8685550093650818,894 "epoch": 0.8436018957345972,895 "grad_norm": 0.7077257037162781,896 "learning_rate": 1.8699063724087905e-05,897 "loss": 0.8259266018867493,898 "mean_token_accuracy": 0.7899300456047058,899 "num_tokens": 1419954.0,900 "step": 89901 },902 {903 "entropy": 0.7893514037132263,904 "epoch": 0.8530805687203792,905 "grad_norm": 0.6737422943115234,906 "learning_rate": 1.866025403784439e-05,907 "loss": 0.7658373713493347,908 "mean_token_accuracy": 0.8061323910951614,909 "num_tokens": 1436833.0,910 "step": 90911 },912 {913 "entropy": 0.874814048409462,914 "epoch": 0.8625592417061612,915 "grad_norm": 0.7337303757667542,916 "learning_rate": 1.862091544856407e-05,917 "loss": 0.881226658821106,918 "mean_token_accuracy": 0.7787951081991196,919 "num_tokens": 1452006.0,920 "step": 91921 },922 {923 "entropy": 0.8939783275127411,924 "epoch": 0.8720379146919431,925 "grad_norm": 0.8276046514511108,926 "learning_rate": 1.8581050358751444e-05,927 "loss": 0.8792139291763306,928 "mean_token_accuracy": 0.7856588512659073,929 "num_tokens": 1467177.0,930 "step": 92931 },932 {933 "entropy": 0.9284755736589432,934 "epoch": 0.8815165876777251,935 "grad_norm": 0.7748722434043884,936 "learning_rate": 1.854066120306571e-05,937 "loss": 0.9296675324440002,938 "mean_token_accuracy": 0.7717541307210922,939 "num_tokens": 1481900.0,940 "step": 93941 },942 {943 "entropy": 0.8540140986442566,944 "epoch": 0.8909952606635071,945 "grad_norm": 0.6757642030715942,946 "learning_rate": 1.8499750448172046e-05,947 "loss": 0.8415413498878479,948 "mean_token_accuracy": 0.7882983386516571,949 "num_tokens": 1498093.0,950 "step": 94951 },952 {953 "entropy": 0.8530154079198837,954 "epoch": 0.9004739336492891,955 "grad_norm": 0.7341352105140686,956 "learning_rate": 1.8458320592590976e-05,957 "loss": 0.8625093698501587,958 "mean_token_accuracy": 0.7759335339069366,959 "num_tokens": 1514287.0,960 "step": 95961 },962 {963 "entropy": 0.7864818871021271,964 "epoch": 0.909952606635071,965 "grad_norm": 0.7490381598472595,966 "learning_rate": 1.841637416654579e-05,967 "loss": 0.8058528900146484,968 "mean_token_accuracy": 0.7964929491281509,969 "num_tokens": 1530706.0,970 "step": 96971 },972 {973 "entropy": 0.8394529819488525,974 "epoch": 0.919431279620853,975 "grad_norm": 0.6680670380592346,976 "learning_rate": 1.837391373180801e-05,977 "loss": 0.8252418041229248,978 "mean_token_accuracy": 0.7954028099775314,979 "num_tokens": 1547595.0,980 "step": 97981 },982 {983 "entropy": 0.8288186490535736,984 "epoch": 0.9289099526066351,985 "grad_norm": 0.7968632578849792,986 "learning_rate": 1.8330941881540917e-05,987 "loss": 0.8349557518959045,988 "mean_token_accuracy": 0.7844376415014267,989 "num_tokens": 1563028.0,990 "step": 98991 },992 {993 "entropy": 0.8333224058151245,994 "epoch": 0.9383886255924171,995 "grad_norm": 0.691230297088623,996 "learning_rate": 1.8287461240141217e-05,997 "loss": 0.8373101353645325,998 "mean_token_accuracy": 0.7931841313838959,999 "num_tokens": 1578461.0,1000 "step": 991001 },1002 {1003 "entropy": 0.9056068807840347,1004 "epoch": 0.9478672985781991,1005 "grad_norm": 0.81464022397995,1006 "learning_rate": 1.8243474463078738e-05,1007 "loss": 0.9318811297416687,1008 "mean_token_accuracy": 0.7732242494821548,1009 "num_tokens": 1594199.0,1010 "step": 1001011 },1012 {1013 "entropy": 0.8589460551738739,1014 "epoch": 0.957345971563981,1015 "grad_norm": 0.8496865630149841,1016 "learning_rate": 1.8198984236734246e-05,1017 "loss": 0.8577089905738831,1018 "mean_token_accuracy": 0.7863648384809494,1019 "num_tokens": 1612009.0,1020 "step": 1011021 },1022 {1023 "entropy": 0.9072522073984146,1024 "epoch": 0.966824644549763,1025 "grad_norm": 0.6795774102210999,1026 "learning_rate": 1.8153993278235416e-05,1027 "loss": 0.9616482853889465,1028 "mean_token_accuracy": 0.7749254554510117,1029 "num_tokens": 1627621.0,1030 "step": 1021031 },1032 {1033 "entropy": 0.824530228972435,1034 "epoch": 0.976303317535545,1035 "grad_norm": 0.7762973308563232,1036 "learning_rate": 1.8108504335290852e-05,1037 "loss": 0.8160134553909302,1038 "mean_token_accuracy": 0.7989892214536667,1039 "num_tokens": 1642633.0,1040 "step": 1031041 },1042 {1043 "entropy": 0.8416652232408524,1044 "epoch": 0.985781990521327,1045 "grad_norm": 0.689736008644104,1046 "learning_rate": 1.80625201860223e-05,1047 "loss": 0.8192683458328247,1048 "mean_token_accuracy": 0.8011076003313065,1049 "num_tokens": 1658656.0,1050 "step": 1041051 },1052 {1053 "entropy": 0.8708516359329224,1054 "epoch": 0.995260663507109,1055 "grad_norm": 0.6898924708366394,1056 "learning_rate": 1.8016043638794975e-05,1057 "loss": 0.8683778047561646,1058 "mean_token_accuracy": 0.788833886384964,1059 "num_tokens": 1674845.0,1060 "step": 1051061 },1062 {1063 "entropy": 0.7872009575366974,1064 "epoch": 1.0,1065 "grad_norm": 0.8945826292037964,1066 "learning_rate": 1.7969077532046047e-05,1067 "loss": 0.7540068030357361,1068 "mean_token_accuracy": 0.8025133311748505,1069 "num_tokens": 1682713.0,1070 "step": 1061071 },1072 {1073 "entropy": 0.8132580518722534,1074 "epoch": 1.009478672985782,1075 "grad_norm": 0.7014424800872803,1076 "learning_rate": 1.7921624734111292e-05,1077 "loss": 0.7927298545837402,1078 "mean_token_accuracy": 0.7970700860023499,1079 "num_tokens": 1698838.0,1080 "step": 1071081 },1082 {1083 "entropy": 0.9552255123853683,1084 "epoch": 1.018957345971564,1085 "grad_norm": 0.7601428031921387,1086 "learning_rate": 1.787368814304992e-05,1087 "loss": 0.9353047609329224,1088 "mean_token_accuracy": 0.7636215835809708,1089 "num_tokens": 1714529.0,1090 "step": 1081091 },1092 {1093 "entropy": 0.8868769556283951,1094 "epoch": 1.028436018957346,1095 "grad_norm": 0.7216587066650391,1096 "learning_rate": 1.7825270686467567e-05,1097 "loss": 0.8972142934799194,1098 "mean_token_accuracy": 0.7766188830137253,1099 "num_tokens": 1730585.0,1100 "step": 1091101 },1102 {1103 "entropy": 0.7498472779989243,1104 "epoch": 1.037914691943128,1105 "grad_norm": 0.7536665201187134,1106 "learning_rate": 1.7776375321337523e-05,1107 "loss": 0.7597789764404297,1108 "mean_token_accuracy": 0.8086548745632172,1109 "num_tokens": 1747998.0,1110 "step": 1101111 },1112 {1113 "entropy": 0.8525176346302032,1114 "epoch": 1.04739336492891,1115 "grad_norm": 0.6725131869316101,1116 "learning_rate": 1.7727005033820117e-05,1117 "loss": 0.8725385665893555,1118 "mean_token_accuracy": 0.7857676595449448,1119 "num_tokens": 1763858.0,1120 "step": 1111121 },1122 {1123 "entropy": 0.8737937361001968,1124 "epoch": 1.0568720379146919,1125 "grad_norm": 0.6200551390647888,1126 "learning_rate": 1.7677162839080365e-05,1127 "loss": 0.8674412965774536,1128 "mean_token_accuracy": 0.7846900969743729,1129 "num_tokens": 1779800.0,1130 "step": 1121131 },1132 {1133 "entropy": 0.8540749698877335,1134 "epoch": 1.066350710900474,1135 "grad_norm": 0.7130475640296936,1136 "learning_rate": 1.762685178110382e-05,1137 "loss": 0.8417346477508545,1138 "mean_token_accuracy": 0.7876130044460297,1139 "num_tokens": 1795033.0,1140 "step": 1131141 },1142 {1143 "entropy": 0.7300510853528976,1144 "epoch": 1.0758293838862558,1145 "grad_norm": 0.7067408561706543,1146 "learning_rate": 1.757607493251067e-05,1147 "loss": 0.7173537015914917,1148 "mean_token_accuracy": 0.8160942941904068,1149 "num_tokens": 1811709.0,1150 "step": 1141151 },1152 {1153 "entropy": 0.7862643301486969,1154 "epoch": 1.085308056872038,1155 "grad_norm": 0.6860957145690918,1156 "learning_rate": 1.752483539436807e-05,1157 "loss": 0.7681137323379517,1158 "mean_token_accuracy": 0.8050192445516586,1159 "num_tokens": 1828031.0,1160 "step": 1151161 },1162 {1163 "entropy": 0.8223290294408798,1164 "epoch": 1.09478672985782,1165 "grad_norm": 13.119089126586914,1166 "learning_rate": 1.747313629600077e-05,1167 "loss": 0.819756805896759,1168 "mean_token_accuracy": 0.7999143302440643,1169 "num_tokens": 1844448.0,1170 "step": 1161171 },1172 {1173 "entropy": 0.8199901729822159,1174 "epoch": 1.1042654028436019,1175 "grad_norm": 0.7526054978370667,1176 "learning_rate": 1.7420980794800013e-05,1177 "loss": 0.8049260377883911,1178 "mean_token_accuracy": 0.7947445511817932,1179 "num_tokens": 1860116.0,1180 "step": 1171181 },1182 {1183 "entropy": 0.9044909924268723,1184 "epoch": 1.113744075829384,1185 "grad_norm": 0.774227499961853,1186 "learning_rate": 1.7368372076030654e-05,1187 "loss": 0.8923323750495911,1188 "mean_token_accuracy": 0.7806955128908157,1189 "num_tokens": 1875432.0,1190 "step": 1181191 },1192 {1193 "entropy": 0.8348246514797211,1194 "epoch": 1.1232227488151658,1195 "grad_norm": 0.767121434211731,1196 "learning_rate": 1.731531335263669e-05,1197 "loss": 0.8475579023361206,1198 "mean_token_accuracy": 0.795254722237587,1199 "num_tokens": 1891887.0,1200 "step": 119