yoinked/simai-codegen
013
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 0.9541984732824428,6 "eval_steps": 500,7 "global_step": 500,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.0019083969465648854,14 "grad_norm": 1.0319172143936157,15 "learning_rate": 0.0002,16 "loss": 1.1888583898544312,17 "step": 118 },19 {20 "epoch": 0.003816793893129771,21 "grad_norm": 2.558675527572632,22 "learning_rate": 0.00019999980030622737,23 "loss": 1.2132498025894165,24 "step": 225 },26 {27 "epoch": 0.0057251908396946565,28 "grad_norm": 1.5326608419418335,29 "learning_rate": 0.00019999920122570704,30 "loss": 1.149939775466919,31 "step": 332 },33 {34 "epoch": 0.007633587786259542,35 "grad_norm": 0.9747706651687622,36 "learning_rate": 0.00019999820276083168,37 "loss": 1.1339305639266968,38 "step": 439 },40 {41 "epoch": 0.009541984732824428,42 "grad_norm": 0.8459669351577759,43 "learning_rate": 0.00019999680491558902,44 "loss": 1.1253714561462402,45 "step": 546 },47 {48 "epoch": 0.011450381679389313,49 "grad_norm": 1.008865237236023,50 "learning_rate": 0.00019999500769556186,51 "loss": 1.2100791931152344,52 "step": 653 },54 {55 "epoch": 0.013358778625954198,56 "grad_norm": 0.8221519589424133,57 "learning_rate": 0.00019999281110792807,58 "loss": 1.2723747491836548,59 "step": 760 },61 {62 "epoch": 0.015267175572519083,63 "grad_norm": 0.8101316094398499,64 "learning_rate": 0.0001999902151614606,65 "loss": 1.1219682693481445,66 "step": 867 },68 {69 "epoch": 0.01717557251908397,70 "grad_norm": 0.9121878743171692,71 "learning_rate": 0.0001999872198665273,72 "loss": 1.1113958358764648,73 "step": 974 },75 {76 "epoch": 0.019083969465648856,77 "grad_norm": 1.0775856971740723,78 "learning_rate": 0.000199983825235091,79 "loss": 1.323795199394226,80 "step": 1081 },82 {83 "epoch": 0.02099236641221374,84 "grad_norm": 0.8592789769172668,85 "learning_rate": 0.00019998003128070946,86 "loss": 1.1396949291229248,87 "step": 1188 },89 {90 "epoch": 0.022900763358778626,91 "grad_norm": 0.7191622257232666,92 "learning_rate": 0.0001999758380185352,93 "loss": 0.872940719127655,94 "step": 1295 },96 {97 "epoch": 0.02480916030534351,98 "grad_norm": 0.8468532562255859,99 "learning_rate": 0.0001999712454653157,100 "loss": 1.0025124549865723,101 "step": 13102 },103 {104 "epoch": 0.026717557251908396,105 "grad_norm": 0.8349865674972534,106 "learning_rate": 0.00019996625363939295,107 "loss": 1.1482839584350586,108 "step": 14109 },110 {111 "epoch": 0.02862595419847328,112 "grad_norm": 0.9650427103042603,113 "learning_rate": 0.0001999608625607037,114 "loss": 1.0744962692260742,115 "step": 15116 },117 {118 "epoch": 0.030534351145038167,119 "grad_norm": 0.766567051410675,120 "learning_rate": 0.00019995507225077926,121 "loss": 1.132323980331421,122 "step": 16123 },124 {125 "epoch": 0.03244274809160305,126 "grad_norm": 0.8034893870353699,127 "learning_rate": 0.0001999488827327454,128 "loss": 0.9454637765884399,129 "step": 17130 },131 {132 "epoch": 0.03435114503816794,133 "grad_norm": 0.9051380157470703,134 "learning_rate": 0.00019994229403132228,135 "loss": 0.8953108191490173,136 "step": 18137 },138 {139 "epoch": 0.03625954198473282,140 "grad_norm": 0.8768414855003357,141 "learning_rate": 0.00019993530617282436,142 "loss": 1.0621753931045532,143 "step": 19144 },145 {146 "epoch": 0.03816793893129771,147 "grad_norm": 0.898906409740448,148 "learning_rate": 0.00019992791918516033,149 "loss": 1.0037579536437988,150 "step": 20151 },152 {153 "epoch": 0.04007633587786259,154 "grad_norm": 0.7816566228866577,155 "learning_rate": 0.00019992013309783277,156 "loss": 0.8949717879295349,157 "step": 21158 },159 {160 "epoch": 0.04198473282442748,161 "grad_norm": 0.7184618711471558,162 "learning_rate": 0.00019991194794193845,163 "loss": 1.0673213005065918,164 "step": 22165 },166 {167 "epoch": 0.04389312977099236,168 "grad_norm": 0.7685574293136597,169 "learning_rate": 0.00019990336375016784,170 "loss": 1.0974361896514893,171 "step": 23172 },173 {174 "epoch": 0.04580152671755725,175 "grad_norm": 0.7797583341598511,176 "learning_rate": 0.0001998943805568051,177 "loss": 0.9971912503242493,178 "step": 24179 },180 {181 "epoch": 0.04770992366412214,182 "grad_norm": 0.6793952584266663,183 "learning_rate": 0.00019988499839772804,184 "loss": 0.9434319138526917,185 "step": 25186 },187 {188 "epoch": 0.04961832061068702,189 "grad_norm": 0.8823168277740479,190 "learning_rate": 0.00019987521731040778,191 "loss": 1.1719576120376587,192 "step": 26193 },194 {195 "epoch": 0.05152671755725191,196 "grad_norm": 0.8843436241149902,197 "learning_rate": 0.00019986503733390883,198 "loss": 1.0748037099838257,199 "step": 27200 },201 {202 "epoch": 0.05343511450381679,203 "grad_norm": 0.6977912187576294,204 "learning_rate": 0.00019985445850888867,205 "loss": 1.0246281623840332,206 "step": 28207 },208 {209 "epoch": 0.05534351145038168,210 "grad_norm": 0.756574273109436,211 "learning_rate": 0.00019984348087759784,212 "loss": 0.9375837445259094,213 "step": 29214 },215 {216 "epoch": 0.05725190839694656,217 "grad_norm": 0.8058716654777527,218 "learning_rate": 0.00019983210448387966,219 "loss": 0.9935528039932251,220 "step": 30221 },222 {223 "epoch": 0.05916030534351145,224 "grad_norm": 0.9702297449111938,225 "learning_rate": 0.00019982032937316998,226 "loss": 1.1681523323059082,227 "step": 31228 },229 {230 "epoch": 0.061068702290076333,231 "grad_norm": 0.8501767516136169,232 "learning_rate": 0.0001998081555924972,233 "loss": 0.9978009462356567,234 "step": 32235 },236 {237 "epoch": 0.06297709923664122,238 "grad_norm": 0.7746052742004395,239 "learning_rate": 0.0001997955831904818,240 "loss": 1.040431022644043,241 "step": 33242 },243 {244 "epoch": 0.0648854961832061,245 "grad_norm": 0.8523740768432617,246 "learning_rate": 0.00019978261221733642,247 "loss": 1.0792338848114014,248 "step": 34249 },250 {251 "epoch": 0.06679389312977099,252 "grad_norm": 0.808559775352478,253 "learning_rate": 0.00019976924272486555,254 "loss": 0.9610211253166199,255 "step": 35256 },257 {258 "epoch": 0.06870229007633588,259 "grad_norm": 0.7221457958221436,260 "learning_rate": 0.00019975547476646524,261 "loss": 0.9176299571990967,262 "step": 36263 },264 {265 "epoch": 0.07061068702290077,266 "grad_norm": 0.680614173412323,267 "learning_rate": 0.000199741308397123,268 "loss": 0.9780462384223938,269 "step": 37270 },271 {272 "epoch": 0.07251908396946564,273 "grad_norm": 1.0235838890075684,274 "learning_rate": 0.00019972674367341756,275 "loss": 1.0874963998794556,276 "step": 38277 },278 {279 "epoch": 0.07442748091603053,280 "grad_norm": 0.8324326276779175,281 "learning_rate": 0.00019971178065351857,282 "loss": 0.9447541236877441,283 "step": 39284 },285 {286 "epoch": 0.07633587786259542,287 "grad_norm": 0.7472124695777893,288 "learning_rate": 0.00019969641939718652,289 "loss": 1.0423003435134888,290 "step": 40291 },292 {293 "epoch": 0.07824427480916031,294 "grad_norm": 0.6840877532958984,295 "learning_rate": 0.00019968065996577234,296 "loss": 1.0449249744415283,297 "step": 41298 },299 {300 "epoch": 0.08015267175572519,301 "grad_norm": 0.8184124827384949,302 "learning_rate": 0.00019966450242221722,303 "loss": 0.9014325141906738,304 "step": 42305 },306 {307 "epoch": 0.08206106870229007,308 "grad_norm": 0.7632555961608887,309 "learning_rate": 0.0001996479468310524,310 "loss": 0.8362308144569397,311 "step": 43312 },313 {314 "epoch": 0.08396946564885496,315 "grad_norm": 0.7152660489082336,316 "learning_rate": 0.00019963099325839883,317 "loss": 0.93896484375,318 "step": 44319 },320 {321 "epoch": 0.08587786259541985,322 "grad_norm": 0.6840109825134277,323 "learning_rate": 0.00019961364177196696,324 "loss": 0.9620580673217773,325 "step": 45326 },327 {328 "epoch": 0.08778625954198473,329 "grad_norm": 0.8080661296844482,330 "learning_rate": 0.0001995958924410565,331 "loss": 0.8377565145492554,332 "step": 46333 },334 {335 "epoch": 0.08969465648854962,336 "grad_norm": 0.7813903093338013,337 "learning_rate": 0.00019957774533655605,338 "loss": 0.8326072096824646,339 "step": 47340 },341 {342 "epoch": 0.0916030534351145,343 "grad_norm": 0.8098936080932617,344 "learning_rate": 0.00019955920053094288,345 "loss": 0.9860023856163025,346 "step": 48347 },348 {349 "epoch": 0.09351145038167939,350 "grad_norm": 0.9210098385810852,351 "learning_rate": 0.00019954025809828266,352 "loss": 0.9747365713119507,353 "step": 49354 },355 {356 "epoch": 0.09541984732824428,357 "grad_norm": 0.7985463738441467,358 "learning_rate": 0.00019952091811422905,359 "loss": 0.9031082987785339,360 "step": 50361 },362 {363 "epoch": 0.09732824427480916,364 "grad_norm": 0.8715912699699402,365 "learning_rate": 0.0001995011806560236,366 "loss": 0.9334609508514404,367 "step": 51368 },369 {370 "epoch": 0.09923664122137404,371 "grad_norm": 0.7367807030677795,372 "learning_rate": 0.0001994810458024952,373 "loss": 0.9366903901100159,374 "step": 52375 },376 {377 "epoch": 0.10114503816793893,378 "grad_norm": 0.8102525472640991,379 "learning_rate": 0.00019946051363405997,380 "loss": 0.9092800617218018,381 "step": 53382 },383 {384 "epoch": 0.10305343511450382,385 "grad_norm": 0.8269142508506775,386 "learning_rate": 0.0001994395842327209,387 "loss": 0.9335564374923706,388 "step": 54389 },390 {391 "epoch": 0.1049618320610687,392 "grad_norm": 0.7004595398902893,393 "learning_rate": 0.0001994182576820673,394 "loss": 0.9463485479354858,395 "step": 55396 },397 {398 "epoch": 0.10687022900763359,399 "grad_norm": 0.9793404936790466,400 "learning_rate": 0.00019939653406727482,401 "loss": 0.9135723114013672,402 "step": 56403 },404 {405 "epoch": 0.10877862595419847,406 "grad_norm": 0.7052667140960693,407 "learning_rate": 0.0001993744134751049,408 "loss": 0.8296093940734863,409 "step": 57410 },411 {412 "epoch": 0.11068702290076336,413 "grad_norm": 0.6771461367607117,414 "learning_rate": 0.00019935189599390435,415 "loss": 0.7838313579559326,416 "step": 58417 },418 {419 "epoch": 0.11259541984732824,420 "grad_norm": 0.7421724200248718,421 "learning_rate": 0.00019932898171360526,422 "loss": 0.9136338233947754,423 "step": 59424 },425 {426 "epoch": 0.11450381679389313,427 "grad_norm": 0.820115327835083,428 "learning_rate": 0.00019930567072572438,429 "loss": 0.9014999866485596,430 "step": 60431 },432 {433 "epoch": 0.11641221374045801,434 "grad_norm": 0.7351306676864624,435 "learning_rate": 0.00019928196312336285,436 "loss": 0.9661044478416443,437 "step": 61438 },439 {440 "epoch": 0.1183206106870229,441 "grad_norm": 0.7182132005691528,442 "learning_rate": 0.00019925785900120596,443 "loss": 0.8359523415565491,444 "step": 62445 },446 {447 "epoch": 0.12022900763358779,448 "grad_norm": 0.630822479724884,449 "learning_rate": 0.0001992333584555225,450 "loss": 0.8223984241485596,451 "step": 63452 },453 {454 "epoch": 0.12213740458015267,455 "grad_norm": 0.8691913485527039,456 "learning_rate": 0.0001992084615841647,457 "loss": 0.9800937175750732,458 "step": 64459 },460 {461 "epoch": 0.12404580152671756,462 "grad_norm": 0.7840291261672974,463 "learning_rate": 0.00019918316848656744,464 "loss": 0.9353207349777222,465 "step": 65466 },467 {468 "epoch": 0.12595419847328243,469 "grad_norm": 0.7878021001815796,470 "learning_rate": 0.00019915747926374828,471 "loss": 0.9968975186347961,472 "step": 66473 },474 {475 "epoch": 0.12786259541984732,476 "grad_norm": 0.7407828569412231,477 "learning_rate": 0.00019913139401830674,478 "loss": 0.947467565536499,479 "step": 67480 },481 {482 "epoch": 0.1297709923664122,483 "grad_norm": 0.7297830581665039,484 "learning_rate": 0.00019910491285442407,485 "loss": 0.9409926533699036,486 "step": 68487 },488 {489 "epoch": 0.1316793893129771,490 "grad_norm": 0.7517293095588684,491 "learning_rate": 0.00019907803587786274,492 "loss": 0.7961177825927734,493 "step": 69494 },495 {496 "epoch": 0.13358778625954199,497 "grad_norm": 0.6893132328987122,498 "learning_rate": 0.000199050763195966,499 "loss": 0.9285712242126465,500 "step": 70501 },502 {503 "epoch": 0.13549618320610687,504 "grad_norm": 0.6858025789260864,505 "learning_rate": 0.0001990230949176576,506 "loss": 0.8860894441604614,507 "step": 71508 },509 {510 "epoch": 0.13740458015267176,511 "grad_norm": 0.9707567691802979,512 "learning_rate": 0.00019899503115344117,513 "loss": 0.8908587694168091,514 "step": 72515 },516 {517 "epoch": 0.13931297709923665,518 "grad_norm": 0.7052493691444397,519 "learning_rate": 0.0001989665720153999,520 "loss": 0.8562502264976501,521 "step": 73522 },523 {524 "epoch": 0.14122137404580154,525 "grad_norm": 0.6578022837638855,526 "learning_rate": 0.00019893771761719605,527 "loss": 0.8883586525917053,528 "step": 74529 },530 {531 "epoch": 0.1431297709923664,532 "grad_norm": 0.7066544890403748,533 "learning_rate": 0.00019890846807407044,534 "loss": 0.976402997970581,535 "step": 75536 },537 {538 "epoch": 0.1450381679389313,539 "grad_norm": 0.9122876524925232,540 "learning_rate": 0.00019887882350284216,541 "loss": 0.9926351308822632,542 "step": 76543 },544 {545 "epoch": 0.14694656488549618,546 "grad_norm": 0.7439401745796204,547 "learning_rate": 0.00019884878402190795,548 "loss": 0.829197108745575,549 "step": 77550 },551 {552 "epoch": 0.14885496183206107,553 "grad_norm": 0.7522874474525452,554 "learning_rate": 0.0001988183497512417,555 "loss": 0.863883912563324,556 "step": 78557 },558 {559 "epoch": 0.15076335877862596,560 "grad_norm": 1.0040572881698608,561 "learning_rate": 0.0001987875208123941,562 "loss": 1.069295883178711,563 "step": 79564 },565 {566 "epoch": 0.15267175572519084,567 "grad_norm": 0.713905930519104,568 "learning_rate": 0.00019875629732849214,569 "loss": 0.8455244898796082,570 "step": 80571 },572 {573 "epoch": 0.15458015267175573,574 "grad_norm": 0.8433374166488647,575 "learning_rate": 0.00019872467942423853,576 "loss": 0.8311648368835449,577 "step": 81578 },579 {580 "epoch": 0.15648854961832062,581 "grad_norm": 0.7732320427894592,582 "learning_rate": 0.00019869266722591115,583 "loss": 0.9065580368041992,584 "step": 82585 },586 {587 "epoch": 0.15839694656488548,588 "grad_norm": 0.7662688493728638,589 "learning_rate": 0.0001986602608613628,590 "loss": 0.9555010795593262,591 "step": 83592 },593 {594 "epoch": 0.16030534351145037,595 "grad_norm": 0.9863274693489075,596 "learning_rate": 0.00019862746046002048,597 "loss": 0.9811774492263794,598 "step": 84599 },600 {601 "epoch": 0.16221374045801526,602 "grad_norm": 0.673652708530426,603 "learning_rate": 0.00019859426615288488,604 "loss": 0.926813006401062,605 "step": 85606 },607 {608 "epoch": 0.16412213740458015,609 "grad_norm": 0.8610010147094727,610 "learning_rate": 0.00019856067807252992,611 "loss": 0.7706037759780884,612 "step": 86613 },614 {615 "epoch": 0.16603053435114504,616 "grad_norm": 0.869065523147583,617 "learning_rate": 0.00019852669635310222,618 "loss": 0.9335610270500183,619 "step": 87620 },621 {622 "epoch": 0.16793893129770993,623 "grad_norm": 0.7144750356674194,624 "learning_rate": 0.00019849232113032053,625 "loss": 0.8826386332511902,626 "step": 88627 },628 {629 "epoch": 0.16984732824427481,630 "grad_norm": 0.6938415765762329,631 "learning_rate": 0.00019845755254147525,632 "loss": 0.8689931631088257,633 "step": 89634 },635 {636 "epoch": 0.1717557251908397,637 "grad_norm": 0.6958433389663696,638 "learning_rate": 0.00019842239072542775,639 "loss": 0.8018342852592468,640 "step": 90641 },642 {643 "epoch": 0.1736641221374046,644 "grad_norm": 0.7545017600059509,645 "learning_rate": 0.00019838683582260993,646 "loss": 0.9315579533576965,647 "step": 91648 },649 {650 "epoch": 0.17557251908396945,651 "grad_norm": 0.8562167882919312,652 "learning_rate": 0.00019835088797502367,653 "loss": 0.950774610042572,654 "step": 92655 },656 {657 "epoch": 0.17748091603053434,658 "grad_norm": 0.7419416904449463,659 "learning_rate": 0.0001983145473262402,660 "loss": 1.0040173530578613,661 "step": 93662 },663 {664 "epoch": 0.17938931297709923,665 "grad_norm": 0.7331036925315857,666 "learning_rate": 0.0001982778140213996,667 "loss": 0.703330934047699,668 "step": 94669 },670 {671 "epoch": 0.18129770992366412,672 "grad_norm": 0.783308207988739,673 "learning_rate": 0.00019824068820720997,674 "loss": 0.9145213961601257,675 "step": 95676 },677 {678 "epoch": 0.183206106870229,679 "grad_norm": 0.7295645475387573,680 "learning_rate": 0.00019820317003194728,681 "loss": 0.8577327728271484,682 "step": 96683 },684 {685 "epoch": 0.1851145038167939,686 "grad_norm": 0.8237174153327942,687 "learning_rate": 0.00019816525964545448,688 "loss": 0.8371306657791138,689 "step": 97690 },691 {692 "epoch": 0.18702290076335878,693 "grad_norm": 0.638324499130249,694 "learning_rate": 0.00019812695719914085,695 "loss": 0.7818671464920044,696 "step": 98697 },698 {699 "epoch": 0.18893129770992367,700 "grad_norm": 0.7308298349380493,701 "learning_rate": 0.00019808826284598165,702 "loss": 0.8242241144180298,703 "step": 99704 },705 {706 "epoch": 0.19083969465648856,707 "grad_norm": 0.8194659948348999,708 "learning_rate": 0.0001980491767405173,709 "loss": 1.0477763414382935,710 "step": 100711 },712 {713 "epoch": 0.19274809160305342,714 "grad_norm": 0.7348308563232422,715 "learning_rate": 0.0001980096990388528,716 "loss": 0.8968179821968079,717 "step": 101718 },719 {720 "epoch": 0.1946564885496183,721 "grad_norm": 0.8615684509277344,722 "learning_rate": 0.00019796982989865722,723 "loss": 0.9265606999397278,724 "step": 102725 },726 {727 "epoch": 0.1965648854961832,728 "grad_norm": 0.9783602952957153,729 "learning_rate": 0.00019792956947916292,730 "loss": 1.0029759407043457,731 "step": 103732 },733 {734 "epoch": 0.1984732824427481,735 "grad_norm": 0.8230413794517517,736 "learning_rate": 0.000197888917941165,737 "loss": 0.9710904955863953,738 "step": 104739 },740 {741 "epoch": 0.20038167938931298,742 "grad_norm": 0.715769350528717,743 "learning_rate": 0.00019784787544702064,744 "loss": 1.0048372745513916,745 "step": 105746 },747 {748 "epoch": 0.20229007633587787,749 "grad_norm": 0.657476544380188,750 "learning_rate": 0.00019780644216064845,751 "loss": 0.9512552618980408,752 "step": 106753 },754 {755 "epoch": 0.20419847328244276,756 "grad_norm": 0.8867478370666504,757 "learning_rate": 0.0001977646182475278,758 "loss": 0.8841535449028015,759 "step": 107760 },761 {762 "epoch": 0.20610687022900764,763 "grad_norm": 0.6844662427902222,764 "learning_rate": 0.00019772240387469823,765 "loss": 0.8210901618003845,766 "step": 108767 },768 {769 "epoch": 0.20801526717557253,770 "grad_norm": 0.8655828833580017,771 "learning_rate": 0.00019767979921075866,772 "loss": 1.0782965421676636,773 "step": 109774 },775 {776 "epoch": 0.2099236641221374,777 "grad_norm": 0.7254316210746765,778 "learning_rate": 0.00019763680442586682,779 "loss": 0.7998471260070801,780 "step": 110781 },782 {783 "epoch": 0.21183206106870228,784 "grad_norm": 0.7396829724311829,785 "learning_rate": 0.00019759341969173853,786 "loss": 0.8773504495620728,787 "step": 111788 },789 {790 "epoch": 0.21374045801526717,791 "grad_norm": 0.6355534791946411,792 "learning_rate": 0.00019754964518164697,793 "loss": 0.8282820582389832,794 "step": 112795 },796 {797 "epoch": 0.21564885496183206,798 "grad_norm": 0.7496469020843506,799 "learning_rate": 0.00019750548107042213,800 "loss": 0.8907659649848938,801 "step": 113802 },803 {804 "epoch": 0.21755725190839695,805 "grad_norm": 0.9808054566383362,806 "learning_rate": 0.00019746092753444999,807 "loss": 0.8398576974868774,808 "step": 114809 },810 {811 "epoch": 0.21946564885496184,812 "grad_norm": 0.8305889964103699,813 "learning_rate": 0.00019741598475167175,814 "loss": 0.9800796508789062,815 "step": 115816 },817 {818 "epoch": 0.22137404580152673,819 "grad_norm": 0.8272994160652161,820 "learning_rate": 0.00019737065290158335,821 "loss": 1.0569555759429932,822 "step": 116823 },824 {825 "epoch": 0.22328244274809161,826 "grad_norm": 0.6429803371429443,827 "learning_rate": 0.00019732493216523448,828 "loss": 0.8179414868354797,829 "step": 117830 },831 {832 "epoch": 0.22519083969465647,833 "grad_norm": 0.7551900148391724,834 "learning_rate": 0.00019727882272522817,835 "loss": 1.0400655269622803,836 "step": 118837 },838 {839 "epoch": 0.22709923664122136,840 "grad_norm": 0.668594479560852,841 "learning_rate": 0.00019723232476571973,842 "loss": 0.7926125526428223,843 "step": 119844 },845 {846 "epoch": 0.22900763358778625,847 "grad_norm": 0.7308197617530823,848 "learning_rate": 0.0001971854384724162,849 "loss": 0.7763074040412903,850 "step": 120851 },852 {853 "epoch": 0.23091603053435114,854 "grad_norm": 0.6899462342262268,855 "learning_rate": 0.0001971381640325756,856 "loss": 0.8050013184547424,857 "step": 121858 },859 {860 "epoch": 0.23282442748091603,861 "grad_norm": 0.6338201761245728,862 "learning_rate": 0.0001970905016350062,863 "loss": 0.7711692452430725,864 "step": 122865 },866 {867 "epoch": 0.23473282442748092,868 "grad_norm": 0.818661093711853,869 "learning_rate": 0.00019704245147006563,870 "loss": 0.9701591730117798,871 "step": 123872 },873 {874 "epoch": 0.2366412213740458,875 "grad_norm": 0.8936249613761902,876 "learning_rate": 0.0001969940137296603,877 "loss": 0.932243287563324,878 "step": 124879 },880 {881 "epoch": 0.2385496183206107,882 "grad_norm": 0.6683034896850586,883 "learning_rate": 0.00019694518860724445,884 "loss": 0.8161548376083374,885 "step": 125886 },887 {888 "epoch": 0.24045801526717558,889 "grad_norm": 0.8327929973602295,890 "learning_rate": 0.00019689597629781962,891 "loss": 0.8570858240127563,892 "step": 126893 },894 {895 "epoch": 0.24236641221374045,896 "grad_norm": 0.8678820729255676,897 "learning_rate": 0.00019684637699793358,898 "loss": 0.849490761756897,899 "step": 127900 },901 {902 "epoch": 0.24427480916030533,903 "grad_norm": 0.8075437545776367,904 "learning_rate": 0.00019679639090567982,905 "loss": 0.7586680054664612,906 "step": 128907 },908 {909 "epoch": 0.24618320610687022,910 "grad_norm": 0.920081377029419,911 "learning_rate": 0.00019674601822069647,912 "loss": 0.9018468260765076,913 "step": 129914 },915 {916 "epoch": 0.2480916030534351,917 "grad_norm": 0.809676468372345,918 "learning_rate": 0.00019669525914416586,919 "loss": 0.9085022807121277,920 "step": 130921 },922 {923 "epoch": 0.25,924 "grad_norm": 0.6830601096153259,925 "learning_rate": 0.00019664411387881336,926 "loss": 0.8158378005027771,927 "step": 131928 },929 {930 "epoch": 0.25190839694656486,931 "grad_norm": 0.8688009977340698,932 "learning_rate": 0.00019659258262890683,933 "loss": 0.8939949870109558,934 "step": 132935 },936 {937 "epoch": 0.2538167938931298,938 "grad_norm": 0.8238283395767212,939 "learning_rate": 0.00019654066560025567,940 "loss": 0.8714411854743958,941 "step": 133942 },943 {944 "epoch": 0.25572519083969464,945 "grad_norm": 0.8402740955352783,946 "learning_rate": 0.00019648836300020995,947 "loss": 0.8553993105888367,948 "step": 134949 },950 {951 "epoch": 0.25763358778625955,952 "grad_norm": 0.8889774680137634,953 "learning_rate": 0.0001964356750376598,954 "loss": 0.9362643361091614,955 "step": 135956 },957 {958 "epoch": 0.2595419847328244,959 "grad_norm": 0.7945830225944519,960 "learning_rate": 0.00019638260192303438,961 "loss": 0.915956437587738,962 "step": 136963 },964 {965 "epoch": 0.26145038167938933,966 "grad_norm": 0.6740367412567139,967 "learning_rate": 0.00019632914386830109,968 "loss": 0.9191530346870422,969 "step": 137970 },971 {972 "epoch": 0.2633587786259542,973 "grad_norm": 0.8426705002784729,974 "learning_rate": 0.00019627530108696473,975 "loss": 0.868916392326355,976 "step": 138977 },978 {979 "epoch": 0.2652671755725191,980 "grad_norm": 0.8657882213592529,981 "learning_rate": 0.00019622107379406667,982 "loss": 0.872461199760437,983 "step": 139984 },985 {986 "epoch": 0.26717557251908397,987 "grad_norm": 0.7300332188606262,988 "learning_rate": 0.00019616646220618398,989 "loss": 0.8058186173439026,990 "step": 140991 },992 {993 "epoch": 0.26908396946564883,994 "grad_norm": 0.767820417881012,995 "learning_rate": 0.00019611146654142854,996 "loss": 0.8053113222122192,997 "step": 141998 },999 {1000 "epoch": 0.27099236641221375,1001 "grad_norm": 0.9955086708068848,1002 "learning_rate": 0.00019605608701944615,1003 "loss": 0.8747786283493042,1004 "step": 1421005 },1006 {1007 "epoch": 0.2729007633587786,1008 "grad_norm": 0.8415116667747498,1009 "learning_rate": 0.0001960003238614158,1010 "loss": 0.7746783494949341,1011 "step": 1431012 },1013 {1014 "epoch": 0.2748091603053435,1015 "grad_norm": 0.7966413497924805,1016 "learning_rate": 0.00019594417729004845,1017 "loss": 0.9076138734817505,1018 "step": 1441019 },1020 {1021 "epoch": 0.2767175572519084,1022 "grad_norm": 0.7794169187545776,1023 "learning_rate": 0.00019588764752958668,1024 "loss": 0.8441069722175598,1025 "step": 1451026 },1027 {1028 "epoch": 0.2786259541984733,1029 "grad_norm": 0.8923323750495911,1030 "learning_rate": 0.00019583073480580323,1031 "loss": 0.9479147791862488,1032 "step": 1461033 },1034 {1035 "epoch": 0.28053435114503816,1036 "grad_norm": 0.7435654401779175,1037 "learning_rate": 0.00019577343934600044,1038 "loss": 0.7812725901603699,1039 "step": 1471040 },1041 {1042 "epoch": 0.2824427480916031,1043 "grad_norm": 0.6814793944358826,1044 "learning_rate": 0.00019571576137900927,1045 "loss": 0.7864100337028503,1046 "step": 1481047 },1048 {1049 "epoch": 0.28435114503816794,1050 "grad_norm": 0.7200238704681396,1051 "learning_rate": 0.00019565770113518825,1052 "loss": 0.9560705423355103,1053 "step": 1491054 },1055 {1056 "epoch": 0.2862595419847328,1057 "grad_norm": 0.6995819807052612,1058 "learning_rate": 0.00019559925884642286,1059 "loss": 0.8137438297271729,1060 "step": 1501061 },1062 {1063 "epoch": 0.2881679389312977,1064 "grad_norm": 0.5875421762466431,1065 "learning_rate": 0.0001955404347461243,1066 "loss": 0.7266908288002014,1067 "step": 1511068 },1069 {1070 "epoch": 0.2900763358778626,1071 "grad_norm": 0.7169772386550903,1072 "learning_rate": 0.00019548122906922865,1073 "loss": 0.8415418863296509,1074 "step": 1521075 },1076 {1077 "epoch": 0.2919847328244275,1078 "grad_norm": 0.638949990272522,1079 "learning_rate": 0.00019542164205219606,1080 "loss": 0.8274745941162109,1081 "step": 1531082 },1083 {1084 "epoch": 0.29389312977099236,1085 "grad_norm": 0.7600482106208801,1086 "learning_rate": 0.00019536167393300965,1087 "loss": 1.0291582345962524,1088 "step": 1541089 },1090 {1091 "epoch": 0.2958015267175573,1092 "grad_norm": 0.7919864058494568,1093 "learning_rate": 0.0001953013249511746,1094 "loss": 0.8221685290336609,1095 "step": 1551096 },1097 {1098 "epoch": 0.29770992366412213,1099 "grad_norm": 0.7551295757293701,1100 "learning_rate": 0.0001952405953477172,1101 "loss": 0.9822176098823547,1102 "step": 1561103 },1104 {1105 "epoch": 0.29961832061068705,1106 "grad_norm": 0.6301788091659546,1107 "learning_rate": 0.000195179485365184,1108 "loss": 0.6514051556587219,1109 "step": 1571110 },1111 {1112 "epoch": 0.3015267175572519,1113 "grad_norm": 0.6198784112930298,1114 "learning_rate": 0.00019511799524764062,1115 "loss": 0.8909632563591003,1116 "step": 1581117 },1118 {1119 "epoch": 0.30343511450381677,1120 "grad_norm": 0.6850892901420593,1121 "learning_rate": 0.0001950561252406709,1122 "loss": 0.8530802726745605,1123 "step": 1591124 },1125 {1126 "epoch": 0.3053435114503817,1127 "grad_norm": 0.6627747416496277,1128 "learning_rate": 0.00019499387559137596,1129 "loss": 0.8950958251953125,1130 "step": 1601131 },1132 {1133 "epoch": 0.30725190839694655,1134 "grad_norm": 0.6648580431938171,1135 "learning_rate": 0.00019493124654837318,1136 "loss": 0.8956643342971802,1137 "step": 1611138 },1139 {1140 "epoch": 0.30916030534351147,1141 "grad_norm": 0.8410931825637817,1142 "learning_rate": 0.00019486823836179512,1143 "loss": 0.8377202153205872,1144 "step": 1621145 },1146 {1147 "epoch": 0.3110687022900763,1148 "grad_norm": 0.7048870921134949,1149 "learning_rate": 0.00019480485128328868,1150 "loss": 0.8189906477928162,1151 "step": 1631152 },1153 {1154 "epoch": 0.31297709923664124,1155 "grad_norm": 0.7690779566764832,1156 "learning_rate": 0.0001947410855660139,1157 "loss": 0.7249743938446045,1158 "step": 1641159 },1160 {1161 "epoch": 0.3148854961832061,1162 "grad_norm": 0.9097875356674194,1163 "learning_rate": 0.0001946769414646431,1164 "loss": 0.9130119681358337,1165 "step": 1651166 },1167 {1168 "epoch": 0.31679389312977096,1169 "grad_norm": 0.798412024974823,1170 "learning_rate": 0.0001946124192353599,1171 "loss": 0.960759699344635,1172 "step": 1661173 },1174 {1175 "epoch": 0.3187022900763359,1176 "grad_norm": 0.6766208410263062,1177 "learning_rate": 0.000194547519135858,1178 "loss": 0.7708785533905029,1179 "step": 1671180 },1181 {1182 "epoch": 0.32061068702290074,1183 "grad_norm": 0.7984035015106201,1184 "learning_rate": 0.00019448224142534033,1185 "loss": 0.9001744985580444,1186 "step": 1681187 },1188 {1189 "epoch": 0.32251908396946566,1190 "grad_norm": 0.7736956477165222,1191 "learning_rate": 0.00019441658636451794,1192 "loss": 0.7067661285400391,1193 "step": 1691194 },1195 {1196 "epoch": 0.3244274809160305,1197 "grad_norm": 0.701378583908081,1198 "learning_rate": 0.00019435055421560896,1199 "loss": 0.8072033524513245,1200 "step": 170