terry69/feedback_p0.1_seed42_level2_syntaxcleanbatch16_reverse
07
1{2 "best_metric": null,3 "best_model_checkpoint": null,4 "epoch": 1.0,5 "eval_steps": 500,6 "global_step": 1863,7 "is_hyper_param_search": false,8 "is_local_process_zero": true,9 "is_world_process_zero": true,10 "log_history": [11 {12 "epoch": 0.0005367686527106817,13 "grad_norm": 24.305655426812773,14 "learning_rate": 5.3475935828877005e-08,15 "loss": 1.3555,16 "step": 117 },18 {19 "epoch": 0.0026838432635534087,20 "grad_norm": 23.22138878044751,21 "learning_rate": 2.6737967914438503e-07,22 "loss": 1.3448,23 "step": 524 },25 {26 "epoch": 0.005367686527106817,27 "grad_norm": 15.39412500867573,28 "learning_rate": 5.347593582887701e-07,29 "loss": 1.3152,30 "step": 1031 },32 {33 "epoch": 0.008051529790660225,34 "grad_norm": 11.814690159057262,35 "learning_rate": 8.021390374331551e-07,36 "loss": 1.1495,37 "step": 1538 },39 {40 "epoch": 0.010735373054213635,41 "grad_norm": 9.400955813538365,42 "learning_rate": 1.0695187165775401e-06,43 "loss": 1.0489,44 "step": 2045 },46 {47 "epoch": 0.013419216317767043,48 "grad_norm": 3.612620121520767,49 "learning_rate": 1.3368983957219254e-06,50 "loss": 0.962,51 "step": 2552 },53 {54 "epoch": 0.01610305958132045,55 "grad_norm": 3.507034366901718,56 "learning_rate": 1.6042780748663103e-06,57 "loss": 0.9095,58 "step": 3059 },60 {61 "epoch": 0.01878690284487386,62 "grad_norm": 2.9769449976377174,63 "learning_rate": 1.8716577540106954e-06,64 "loss": 0.8651,65 "step": 3566 },67 {68 "epoch": 0.02147074610842727,69 "grad_norm": 2.963828184373544,70 "learning_rate": 2.1390374331550802e-06,71 "loss": 0.8482,72 "step": 4073 },74 {75 "epoch": 0.024154589371980676,76 "grad_norm": 3.1133147768329232,77 "learning_rate": 2.4064171122994653e-06,78 "loss": 0.85,79 "step": 4580 },81 {82 "epoch": 0.026838432635534086,83 "grad_norm": 2.931545369425598,84 "learning_rate": 2.673796791443851e-06,85 "loss": 0.8309,86 "step": 5087 },88 {89 "epoch": 0.029522275899087493,90 "grad_norm": 3.0256227773094384,91 "learning_rate": 2.9411764705882355e-06,92 "loss": 0.8186,93 "step": 5594 },95 {96 "epoch": 0.0322061191626409,97 "grad_norm": 3.0925781868322955,98 "learning_rate": 3.2085561497326205e-06,99 "loss": 0.8041,100 "step": 60101 },102 {103 "epoch": 0.03488996242619431,104 "grad_norm": 3.063930015490271,105 "learning_rate": 3.4759358288770056e-06,106 "loss": 0.806,107 "step": 65108 },109 {110 "epoch": 0.03757380568974772,111 "grad_norm": 3.2043146977534662,112 "learning_rate": 3.7433155080213907e-06,113 "loss": 0.8075,114 "step": 70115 },116 {117 "epoch": 0.040257648953301126,118 "grad_norm": 2.92763825791418,119 "learning_rate": 4.010695187165775e-06,120 "loss": 0.7952,121 "step": 75122 },123 {124 "epoch": 0.04294149221685454,125 "grad_norm": 3.0927861281954665,126 "learning_rate": 4.2780748663101604e-06,127 "loss": 0.7924,128 "step": 80129 },130 {131 "epoch": 0.045625335480407946,132 "grad_norm": 3.0187031080233817,133 "learning_rate": 4.5454545454545455e-06,134 "loss": 0.7472,135 "step": 85136 },137 {138 "epoch": 0.04830917874396135,139 "grad_norm": 2.974576768839193,140 "learning_rate": 4.812834224598931e-06,141 "loss": 0.7615,142 "step": 90143 },144 {145 "epoch": 0.05099302200751476,146 "grad_norm": 2.986145546612386,147 "learning_rate": 5.0802139037433165e-06,148 "loss": 0.7619,149 "step": 95150 },151 {152 "epoch": 0.05367686527106817,153 "grad_norm": 3.0743318362981755,154 "learning_rate": 5.347593582887702e-06,155 "loss": 0.7617,156 "step": 100157 },158 {159 "epoch": 0.05636070853462158,160 "grad_norm": 3.2918058213261094,161 "learning_rate": 5.614973262032086e-06,162 "loss": 0.7675,163 "step": 105164 },165 {166 "epoch": 0.059044551798174985,167 "grad_norm": 3.2237072710532493,168 "learning_rate": 5.882352941176471e-06,169 "loss": 0.7331,170 "step": 110171 },172 {173 "epoch": 0.06172839506172839,174 "grad_norm": 3.1899890654845606,175 "learning_rate": 6.149732620320856e-06,176 "loss": 0.7459,177 "step": 115178 },179 {180 "epoch": 0.0644122383252818,181 "grad_norm": 3.043596246389249,182 "learning_rate": 6.417112299465241e-06,183 "loss": 0.744,184 "step": 120185 },186 {187 "epoch": 0.06709608158883522,188 "grad_norm": 2.9018863887175743,189 "learning_rate": 6.684491978609626e-06,190 "loss": 0.7377,191 "step": 125192 },193 {194 "epoch": 0.06977992485238863,195 "grad_norm": 2.907463722040111,196 "learning_rate": 6.951871657754011e-06,197 "loss": 0.7427,198 "step": 130199 },200 {201 "epoch": 0.07246376811594203,202 "grad_norm": 2.84315815544186,203 "learning_rate": 7.219251336898396e-06,204 "loss": 0.7319,205 "step": 135206 },207 {208 "epoch": 0.07514761137949544,209 "grad_norm": 2.973654603626244,210 "learning_rate": 7.486631016042781e-06,211 "loss": 0.7328,212 "step": 140213 },214 {215 "epoch": 0.07783145464304884,216 "grad_norm": 3.141459628944732,217 "learning_rate": 7.754010695187166e-06,218 "loss": 0.7338,219 "step": 145220 },221 {222 "epoch": 0.08051529790660225,223 "grad_norm": 2.842095254666257,224 "learning_rate": 8.02139037433155e-06,225 "loss": 0.7343,226 "step": 150227 },228 {229 "epoch": 0.08319914117015566,230 "grad_norm": 2.868583490076937,231 "learning_rate": 8.288770053475937e-06,232 "loss": 0.7287,233 "step": 155234 },235 {236 "epoch": 0.08588298443370908,237 "grad_norm": 2.8470718739082255,238 "learning_rate": 8.556149732620321e-06,239 "loss": 0.7195,240 "step": 160241 },242 {243 "epoch": 0.08856682769726248,244 "grad_norm": 2.9010358259004065,245 "learning_rate": 8.823529411764707e-06,246 "loss": 0.7036,247 "step": 165248 },249 {250 "epoch": 0.09125067096081589,251 "grad_norm": 3.305646484034517,252 "learning_rate": 9.090909090909091e-06,253 "loss": 0.7215,254 "step": 170255 },256 {257 "epoch": 0.0939345142243693,258 "grad_norm": 2.969331866099511,259 "learning_rate": 9.358288770053477e-06,260 "loss": 0.721,261 "step": 175262 },263 {264 "epoch": 0.0966183574879227,265 "grad_norm": 2.93754880819952,266 "learning_rate": 9.625668449197861e-06,267 "loss": 0.7038,268 "step": 180269 },270 {271 "epoch": 0.09930220075147611,272 "grad_norm": 2.7050202580005065,273 "learning_rate": 9.893048128342247e-06,274 "loss": 0.7206,275 "step": 185276 },277 {278 "epoch": 0.10198604401502952,279 "grad_norm": 3.047634989404141,280 "learning_rate": 9.999920944317846e-06,281 "loss": 0.6872,282 "step": 190283 },284 {285 "epoch": 0.10466988727858294,286 "grad_norm": 2.795885504909657,287 "learning_rate": 9.99943783531341e-06,288 "loss": 0.7127,289 "step": 195290 },291 {292 "epoch": 0.10735373054213634,293 "grad_norm": 2.744750472731151,294 "learning_rate": 9.998515579512446e-06,295 "loss": 0.7165,296 "step": 200297 },298 {299 "epoch": 0.11003757380568975,300 "grad_norm": 2.7651055012729104,301 "learning_rate": 9.997154257925199e-06,302 "loss": 0.7171,303 "step": 205304 },305 {306 "epoch": 0.11272141706924316,307 "grad_norm": 3.206532585503099,308 "learning_rate": 9.995353990129115e-06,309 "loss": 0.695,310 "step": 210311 },312 {313 "epoch": 0.11540526033279656,314 "grad_norm": 3.024589614754047,315 "learning_rate": 9.99311493425834e-06,316 "loss": 0.7037,317 "step": 215318 },319 {320 "epoch": 0.11808910359634997,321 "grad_norm": 2.551935026870453,322 "learning_rate": 9.99043728698983e-06,323 "loss": 0.6877,324 "step": 220325 },326 {327 "epoch": 0.12077294685990338,328 "grad_norm": 2.539435880635434,329 "learning_rate": 9.987321283526072e-06,330 "loss": 0.7087,331 "step": 225332 },333 {334 "epoch": 0.12345679012345678,335 "grad_norm": 2.676637809439421,336 "learning_rate": 9.983767197574432e-06,337 "loss": 0.6936,338 "step": 230339 },340 {341 "epoch": 0.1261406333870102,342 "grad_norm": 2.7492296664847045,343 "learning_rate": 9.979775341323097e-06,344 "loss": 0.6915,345 "step": 235346 },347 {348 "epoch": 0.1288244766505636,349 "grad_norm": 2.6934789882553742,350 "learning_rate": 9.975346065413673e-06,351 "loss": 0.7031,352 "step": 240353 },354 {355 "epoch": 0.13150831991411702,356 "grad_norm": 2.640702210698053,357 "learning_rate": 9.970479758910365e-06,358 "loss": 0.7108,359 "step": 245360 },361 {362 "epoch": 0.13419216317767044,363 "grad_norm": 2.748234011053248,364 "learning_rate": 9.965176849265814e-06,365 "loss": 0.7027,366 "step": 250367 },368 {369 "epoch": 0.13687600644122383,370 "grad_norm": 2.5844501833121294,371 "learning_rate": 9.959437802283552e-06,372 "loss": 0.6878,373 "step": 255374 },375 {376 "epoch": 0.13955984970477725,377 "grad_norm": 2.66470810089039,378 "learning_rate": 9.953263122077077e-06,379 "loss": 0.6805,380 "step": 260381 },382 {383 "epoch": 0.14224369296833064,384 "grad_norm": 2.4613007519929866,385 "learning_rate": 9.946653351025575e-06,386 "loss": 0.6808,387 "step": 265388 },389 {390 "epoch": 0.14492753623188406,391 "grad_norm": 2.7303364431033583,392 "learning_rate": 9.939609069726279e-06,393 "loss": 0.6872,394 "step": 270395 },396 {397 "epoch": 0.14761137949543746,398 "grad_norm": 2.4095297951330865,399 "learning_rate": 9.932130896943477e-06,400 "loss": 0.6788,401 "step": 275402 },403 {404 "epoch": 0.15029522275899088,405 "grad_norm": 2.5247027802281847,406 "learning_rate": 9.924219489554145e-06,407 "loss": 0.6903,408 "step": 280409 },410 {411 "epoch": 0.1529790660225443,412 "grad_norm": 2.48852250288296,413 "learning_rate": 9.915875542490257e-06,414 "loss": 0.6873,415 "step": 285416 },417 {418 "epoch": 0.1556629092860977,419 "grad_norm": 2.5324167659849333,420 "learning_rate": 9.907099788677745e-06,421 "loss": 0.6576,422 "step": 290423 },424 {425 "epoch": 0.1583467525496511,426 "grad_norm": 2.564293700320395,427 "learning_rate": 9.897892998972113e-06,428 "loss": 0.6685,429 "step": 295430 },431 {432 "epoch": 0.1610305958132045,433 "grad_norm": 2.7292649090364183,434 "learning_rate": 9.888255982090728e-06,435 "loss": 0.6812,436 "step": 300437 },438 {439 "epoch": 0.16371443907675792,440 "grad_norm": 2.43813150224106,441 "learning_rate": 9.878189584541783e-06,442 "loss": 0.6707,443 "step": 305444 },445 {446 "epoch": 0.16639828234031132,447 "grad_norm": 4.145173454039506,448 "learning_rate": 9.867694690549943e-06,449 "loss": 0.6579,450 "step": 310451 },452 {453 "epoch": 0.16908212560386474,454 "grad_norm": 2.5276238351648552,455 "learning_rate": 9.85677222197867e-06,456 "loss": 0.6707,457 "step": 315458 },459 {460 "epoch": 0.17176596886741816,461 "grad_norm": 2.482449160388171,462 "learning_rate": 9.845423138249254e-06,463 "loss": 0.6629,464 "step": 320465 },466 {467 "epoch": 0.17444981213097155,468 "grad_norm": 2.426718943488087,469 "learning_rate": 9.833648436256525e-06,470 "loss": 0.6677,471 "step": 325472 },473 {474 "epoch": 0.17713365539452497,475 "grad_norm": 2.453880759981358,476 "learning_rate": 9.821449150281308e-06,477 "loss": 0.6725,478 "step": 330479 },480 {481 "epoch": 0.17981749865807836,482 "grad_norm": 2.539784755342757,483 "learning_rate": 9.808826351899551e-06,484 "loss": 0.6626,485 "step": 335486 },487 {488 "epoch": 0.18250134192163178,489 "grad_norm": 2.519434594701762,490 "learning_rate": 9.795781149888216e-06,491 "loss": 0.666,492 "step": 340493 },494 {495 "epoch": 0.18518518518518517,496 "grad_norm": 2.7383878655688814,497 "learning_rate": 9.782314690127867e-06,498 "loss": 0.6448,499 "step": 345500 },501 {502 "epoch": 0.1878690284487386,503 "grad_norm": 2.64399762195394,504 "learning_rate": 9.768428155502038e-06,505 "loss": 0.654,506 "step": 350507 },508 {509 "epoch": 0.19055287171229202,510 "grad_norm": 2.4804265805545134,511 "learning_rate": 9.754122765793306e-06,512 "loss": 0.629,513 "step": 355514 },515 {516 "epoch": 0.1932367149758454,517 "grad_norm": 2.543201466486224,518 "learning_rate": 9.739399777576169e-06,519 "loss": 0.6409,520 "step": 360521 },522 {523 "epoch": 0.19592055823939883,524 "grad_norm": 2.5609743466777397,525 "learning_rate": 9.72426048410665e-06,526 "loss": 0.6568,527 "step": 365528 },529 {530 "epoch": 0.19860440150295222,531 "grad_norm": 2.403379282200782,532 "learning_rate": 9.70870621520871e-06,533 "loss": 0.645,534 "step": 370535 },536 {537 "epoch": 0.20128824476650564,538 "grad_norm": 2.3259878238624014,539 "learning_rate": 9.692738337157441e-06,540 "loss": 0.6507,541 "step": 375542 },543 {544 "epoch": 0.20397208803005903,545 "grad_norm": 2.5652621488474936,546 "learning_rate": 9.676358252559034e-06,547 "loss": 0.6441,548 "step": 380549 },550 {551 "epoch": 0.20665593129361245,552 "grad_norm": 2.7173664219726934,553 "learning_rate": 9.6595674002276e-06,554 "loss": 0.6407,555 "step": 385556 },557 {558 "epoch": 0.20933977455716588,559 "grad_norm": 2.6039911875109563,560 "learning_rate": 9.642367255058767e-06,561 "loss": 0.6218,562 "step": 390563 },564 {565 "epoch": 0.21202361782071927,566 "grad_norm": 2.344579879458719,567 "learning_rate": 9.624759327900131e-06,568 "loss": 0.6536,569 "step": 395570 },571 {572 "epoch": 0.2147074610842727,573 "grad_norm": 2.6297589664652947,574 "learning_rate": 9.606745165418554e-06,575 "loss": 0.6363,576 "step": 400577 },578 {579 "epoch": 0.21739130434782608,580 "grad_norm": 2.39785560505727,581 "learning_rate": 9.58832634996429e-06,582 "loss": 0.6436,583 "step": 405584 },585 {586 "epoch": 0.2200751476113795,587 "grad_norm": 2.585642851240206,588 "learning_rate": 9.569504499432005e-06,589 "loss": 0.6258,590 "step": 410591 },592 {593 "epoch": 0.2227589908749329,594 "grad_norm": 2.8691418365076053,595 "learning_rate": 9.550281267118659e-06,596 "loss": 0.6275,597 "step": 415598 },599 {600 "epoch": 0.22544283413848631,601 "grad_norm": 2.476684695004809,602 "learning_rate": 9.530658341578276e-06,603 "loss": 0.6417,604 "step": 420605 },606 {607 "epoch": 0.2281266774020397,608 "grad_norm": 2.363278675306579,609 "learning_rate": 9.510637446473633e-06,610 "loss": 0.6242,611 "step": 425612 },613 {614 "epoch": 0.23081052066559313,615 "grad_norm": 2.4597379363178624,616 "learning_rate": 9.490220340424844e-06,617 "loss": 0.6375,618 "step": 430619 },620 {621 "epoch": 0.23349436392914655,622 "grad_norm": 2.383668683197536,623 "learning_rate": 9.469408816854898e-06,624 "loss": 0.6058,625 "step": 435626 },627 {628 "epoch": 0.23617820719269994,629 "grad_norm": 2.355853732674972,630 "learning_rate": 9.448204703832102e-06,631 "loss": 0.6156,632 "step": 440633 },634 {635 "epoch": 0.23886205045625336,636 "grad_norm": 2.3375352653671646,637 "learning_rate": 9.426609863909537e-06,638 "loss": 0.6238,639 "step": 445640 },641 {642 "epoch": 0.24154589371980675,643 "grad_norm": 2.547634943087699,644 "learning_rate": 9.404626193961427e-06,645 "loss": 0.6186,646 "step": 450647 },648 {649 "epoch": 0.24422973698336017,650 "grad_norm": 2.719745903154331,651 "learning_rate": 9.382255625016527e-06,652 "loss": 0.6312,653 "step": 455654 },655 {656 "epoch": 0.24691358024691357,657 "grad_norm": 2.8633960737686164,658 "learning_rate": 9.359500122088511e-06,659 "loss": 0.6158,660 "step": 460661 },662 {663 "epoch": 0.249597423510467,664 "grad_norm": 2.4972600338410573,665 "learning_rate": 9.336361684003353e-06,666 "loss": 0.629,667 "step": 465668 },669 {670 "epoch": 0.2522812667740204,671 "grad_norm": 2.2735215195501697,672 "learning_rate": 9.312842343223764e-06,673 "loss": 0.6106,674 "step": 470675 },676 {677 "epoch": 0.2549651100375738,678 "grad_norm": 2.365141838824348,679 "learning_rate": 9.288944165670651e-06,680 "loss": 0.6059,681 "step": 475682 },683 {684 "epoch": 0.2576489533011272,685 "grad_norm": 2.5760090652843046,686 "learning_rate": 9.264669250541658e-06,687 "loss": 0.6246,688 "step": 480689 },690 {691 "epoch": 0.2603327965646806,692 "grad_norm": 2.5329655943720226,693 "learning_rate": 9.240019730126764e-06,694 "loss": 0.606,695 "step": 485696 },697 {698 "epoch": 0.26301663982823403,699 "grad_norm": 2.5738935216730003,700 "learning_rate": 9.214997769620998e-06,701 "loss": 0.6143,702 "step": 490703 },704 {705 "epoch": 0.26570048309178745,706 "grad_norm": 2.63987142371882,707 "learning_rate": 9.189605566934235e-06,708 "loss": 0.602,709 "step": 495710 },711 {712 "epoch": 0.2683843263553409,713 "grad_norm": 2.322948855489504,714 "learning_rate": 9.163845352498141e-06,715 "loss": 0.5985,716 "step": 500717 },718 {719 "epoch": 0.27106816961889424,720 "grad_norm": 2.402976520275366,721 "learning_rate": 9.137719389070259e-06,722 "loss": 0.608,723 "step": 505724 },725 {726 "epoch": 0.27375201288244766,727 "grad_norm": 2.456701559661602,728 "learning_rate": 9.111229971535231e-06,729 "loss": 0.6084,730 "step": 510731 },732 {733 "epoch": 0.2764358561460011,734 "grad_norm": 2.396999722003953,735 "learning_rate": 9.084379426703245e-06,736 "loss": 0.6078,737 "step": 515738 },739 {740 "epoch": 0.2791196994095545,741 "grad_norm": 2.2944656877855123,742 "learning_rate": 9.057170113105622e-06,743 "loss": 0.5982,744 "step": 520745 },746 {747 "epoch": 0.28180354267310787,748 "grad_norm": 2.5488002762479773,749 "learning_rate": 9.029604420787666e-06,750 "loss": 0.595,751 "step": 525752 },753 {754 "epoch": 0.2844873859366613,755 "grad_norm": 2.5621814193544656,756 "learning_rate": 9.001684771098709e-06,757 "loss": 0.5849,758 "step": 530759 },760 {761 "epoch": 0.2871712292002147,762 "grad_norm": 2.4681234501326927,763 "learning_rate": 8.973413616479429e-06,764 "loss": 0.5943,765 "step": 535766 },767 {768 "epoch": 0.2898550724637681,769 "grad_norm": 2.722213219660786,770 "learning_rate": 8.944793440246435e-06,771 "loss": 0.6,772 "step": 540773 },774 {775 "epoch": 0.29253891572732155,776 "grad_norm": 2.482994818215236,777 "learning_rate": 8.915826756374118e-06,778 "loss": 0.6043,779 "step": 545780 },781 {782 "epoch": 0.2952227589908749,783 "grad_norm": 2.3130329135558867,784 "learning_rate": 8.88651610927384e-06,785 "loss": 0.585,786 "step": 550787 },788 {789 "epoch": 0.29790660225442833,790 "grad_norm": 2.608504562134939,791 "learning_rate": 8.856864073570429e-06,792 "loss": 0.5769,793 "step": 555794 },795 {796 "epoch": 0.30059044551798175,797 "grad_norm": 2.518035685353769,798 "learning_rate": 8.82687325387603e-06,799 "loss": 0.5684,800 "step": 560801 },802 {803 "epoch": 0.3032742887815352,804 "grad_norm": 2.3907513070004662,805 "learning_rate": 8.796546284561307e-06,806 "loss": 0.5925,807 "step": 565808 },809 {810 "epoch": 0.3059581320450886,811 "grad_norm": 2.3102125436889347,812 "learning_rate": 8.765885829524059e-06,813 "loss": 0.5875,814 "step": 570815 },816 {817 "epoch": 0.30864197530864196,818 "grad_norm": 2.3679548171728224,819 "learning_rate": 8.734894581955208e-06,820 "loss": 0.5935,821 "step": 575822 },823 {824 "epoch": 0.3113258185721954,825 "grad_norm": 2.384663332405018,826 "learning_rate": 8.703575264102245e-06,827 "loss": 0.587,828 "step": 580829 },830 {831 "epoch": 0.3140096618357488,832 "grad_norm": 2.4842753819177177,833 "learning_rate": 8.6719306270301e-06,834 "loss": 0.5877,835 "step": 585836 },837 {838 "epoch": 0.3166935050993022,839 "grad_norm": 2.4604537301515355,840 "learning_rate": 8.639963450379494e-06,841 "loss": 0.5937,842 "step": 590843 },844 {845 "epoch": 0.3193773483628556,846 "grad_norm": 2.3929768937522464,847 "learning_rate": 8.607676542122782e-06,848 "loss": 0.5873,849 "step": 595850 },851 {852 "epoch": 0.322061191626409,853 "grad_norm": 2.2712444923217303,854 "learning_rate": 8.575072738317297e-06,855 "loss": 0.5867,856 "step": 600857 },858 {859 "epoch": 0.3247450348899624,860 "grad_norm": 2.4285841734633054,861 "learning_rate": 8.542154902856232e-06,862 "loss": 0.5798,863 "step": 605864 },865 {866 "epoch": 0.32742887815351585,867 "grad_norm": 2.5643185358166165,868 "learning_rate": 8.508925927217083e-06,869 "loss": 0.5696,870 "step": 610871 },872 {873 "epoch": 0.33011272141706927,874 "grad_norm": 2.480969569279566,875 "learning_rate": 8.475388730207662e-06,876 "loss": 0.5531,877 "step": 615878 },879 {880 "epoch": 0.33279656468062263,881 "grad_norm": 2.4832540196963455,882 "learning_rate": 8.441546257709708e-06,883 "loss": 0.5653,884 "step": 620885 },886 {887 "epoch": 0.33548040794417605,888 "grad_norm": 2.3354177059858623,889 "learning_rate": 8.407401482420129e-06,890 "loss": 0.5788,891 "step": 625892 },893 {894 "epoch": 0.33816425120772947,895 "grad_norm": 2.542916086554623,896 "learning_rate": 8.372957403589873e-06,897 "loss": 0.5626,898 "step": 630899 },900 {901 "epoch": 0.3408480944712829,902 "grad_norm": 2.469264768397056,903 "learning_rate": 8.33821704676049e-06,904 "loss": 0.5641,905 "step": 635906 },907 {908 "epoch": 0.3435319377348363,909 "grad_norm": 2.335932114462721,910 "learning_rate": 8.303183463498357e-06,911 "loss": 0.5588,912 "step": 640913 },914 {915 "epoch": 0.3462157809983897,916 "grad_norm": 2.3426121287532506,917 "learning_rate": 8.26785973112664e-06,918 "loss": 0.5665,919 "step": 645920 },921 {922 "epoch": 0.3488996242619431,923 "grad_norm": 2.5085502164744407,924 "learning_rate": 8.232248952454978e-06,925 "loss": 0.5551,926 "step": 650927 },928 {929 "epoch": 0.3515834675254965,930 "grad_norm": 2.367908978354936,931 "learning_rate": 8.196354255506937e-06,932 "loss": 0.5841,933 "step": 655934 },935 {936 "epoch": 0.35426731078904994,937 "grad_norm": 2.368147432676988,938 "learning_rate": 8.160178793245254e-06,939 "loss": 0.5543,940 "step": 660941 },942 {943 "epoch": 0.3569511540526033,944 "grad_norm": 2.4274434068510464,945 "learning_rate": 8.12372574329487e-06,946 "loss": 0.553,947 "step": 665948 },949 {950 "epoch": 0.3596349973161567,951 "grad_norm": 2.42778146542873,952 "learning_rate": 8.086998307663815e-06,953 "loss": 0.5613,954 "step": 670955 },956 {957 "epoch": 0.36231884057971014,958 "grad_norm": 2.38111212352637,959 "learning_rate": 8.049999712461956e-06,960 "loss": 0.5437,961 "step": 675962 },963 {964 "epoch": 0.36500268384326356,965 "grad_norm": 2.33473536282302,966 "learning_rate": 8.012733207617602e-06,967 "loss": 0.5519,968 "step": 680969 },970 {971 "epoch": 0.367686527106817,972 "grad_norm": 2.260812226426273,973 "learning_rate": 7.975202066592038e-06,974 "loss": 0.5453,975 "step": 685976 },977 {978 "epoch": 0.37037037037037035,979 "grad_norm": 2.463629677104882,980 "learning_rate": 7.937409586091996e-06,981 "loss": 0.552,982 "step": 690983 },984 {985 "epoch": 0.37305421363392377,986 "grad_norm": 2.507311462473454,987 "learning_rate": 7.899359085780062e-06,988 "loss": 0.5482,989 "step": 695990 },991 {992 "epoch": 0.3757380568974772,993 "grad_norm": 2.4632736487546905,994 "learning_rate": 7.86105390798309e-06,995 "loss": 0.5416,996 "step": 700997 },998 {999 "epoch": 0.3784219001610306,1000 "grad_norm": 2.4592304831976333,1001 "learning_rate": 7.822497417398607e-06,1002 "loss": 0.5461,1003 "step": 7051004 },1005 {1006 "epoch": 0.38110574342458403,1007 "grad_norm": 2.3972861209417484,1008 "learning_rate": 7.783693000799263e-06,1009 "loss": 0.5415,1010 "step": 7101011 },1012 {1013 "epoch": 0.3837895866881374,1014 "grad_norm": 2.385175750668056,1015 "learning_rate": 7.744644066735335e-06,1016 "loss": 0.529,1017 "step": 7151018 },1019 {1020 "epoch": 0.3864734299516908,1021 "grad_norm": 2.2941982595733634,1022 "learning_rate": 7.705354045235335e-06,1023 "loss": 0.5263,1024 "step": 7201025 },1026 {1027 "epoch": 0.38915727321524424,1028 "grad_norm": 2.4098039236931217,1029 "learning_rate": 7.665826387504708e-06,1030 "loss": 0.5367,1031 "step": 7251032 },1033 {1034 "epoch": 0.39184111647879766,1035 "grad_norm": 2.369548787017956,1036 "learning_rate": 7.626064565622684e-06,1037 "loss": 0.5492,1038 "step": 7301039 },1040 {1041 "epoch": 0.394524959742351,1042 "grad_norm": 2.485485577179457,1043 "learning_rate": 7.586072072237291e-06,1044 "loss": 0.549,1045 "step": 7351046 },1047 {1048 "epoch": 0.39720880300590444,1049 "grad_norm": 2.3413134338829678,1050 "learning_rate": 7.545852420258566e-06,1051 "loss": 0.5416,1052 "step": 7401053 },1054 {1055 "epoch": 0.39989264626945786,1056 "grad_norm": 2.4643129883436163,1057 "learning_rate": 7.505409142549987e-06,1058 "loss": 0.5317,1059 "step": 7451060 },1061 {1062 "epoch": 0.4025764895330113,1063 "grad_norm": 2.4296220280963796,1064 "learning_rate": 7.464745791618138e-06,1065 "loss": 0.517,1066 "step": 7501067 },1068 {1069 "epoch": 0.4052603327965647,1070 "grad_norm": 2.2974621401486988,1071 "learning_rate": 7.423865939300674e-06,1072 "loss": 0.5291,1073 "step": 7551074 },1075 {1076 "epoch": 0.40794417606011807,1077 "grad_norm": 2.3986283475814685,1078 "learning_rate": 7.382773176452561e-06,1079 "loss": 0.5143,1080 "step": 7601081 },1082 {1083 "epoch": 0.4106280193236715,1084 "grad_norm": 2.4550380944787893,1085 "learning_rate": 7.341471112630667e-06,1086 "loss": 0.5303,1087 "step": 7651088 },1089 {1090 "epoch": 0.4133118625872249,1091 "grad_norm": 2.439286617193026,1092 "learning_rate": 7.2999633757766956e-06,1093 "loss": 0.5209,1094 "step": 7701095 },1096 {1097 "epoch": 0.41599570585077833,1098 "grad_norm": 2.579280240608442,1099 "learning_rate": 7.258253611898509e-06,1100 "loss": 0.5181,1101 "step": 7751102 },1103 {1104 "epoch": 0.41867954911433175,1105 "grad_norm": 2.4459646375459707,1106 "learning_rate": 7.216345484749876e-06,1107 "loss": 0.5285,1108 "step": 7801109 },1110 {1111 "epoch": 0.4213633923778851,1112 "grad_norm": 2.31347148751011,1113 "learning_rate": 7.174242675508636e-06,1114 "loss": 0.5209,1115 "step": 7851116 },1117 {1118 "epoch": 0.42404723564143854,1119 "grad_norm": 2.2544720712727306,1120 "learning_rate": 7.131948882453361e-06,1121 "loss": 0.5189,1122 "step": 7901123 },1124 {1125 "epoch": 0.42673107890499196,1126 "grad_norm": 2.4273603418815353,1127 "learning_rate": 7.089467820638491e-06,1128 "loss": 0.5185,1129 "step": 7951130 },1131 {1132 "epoch": 0.4294149221685454,1133 "grad_norm": 2.474461554762193,1134 "learning_rate": 7.0468032215680116e-06,1135 "loss": 0.5005,1136 "step": 8001137 },1138 {1139 "epoch": 0.43209876543209874,1140 "grad_norm": 2.3040840560842515,1141 "learning_rate": 7.003958832867681e-06,1142 "loss": 0.5089,1143 "step": 8051144 },1145 {1146 "epoch": 0.43478260869565216,1147 "grad_norm": 2.3521717336924923,1148 "learning_rate": 6.960938417955841e-06,1149 "loss": 0.4912,1150 "step": 8101151 },1152 {1153 "epoch": 0.4374664519592056,1154 "grad_norm": 2.645750501936123,1155 "learning_rate": 6.917745755712839e-06,1156 "loss": 0.5103,1157 "step": 8151158 },1159 {1160 "epoch": 0.440150295222759,1161 "grad_norm": 2.3864966496000775,1162 "learning_rate": 6.874384640149101e-06,1163 "loss": 0.5173,1164 "step": 8201165 },1166 {1167 "epoch": 0.4428341384863124,1168 "grad_norm": 2.4243382348421023,1169 "learning_rate": 6.830858880071859e-06,1170 "loss": 0.5008,1171 "step": 8251172 },1173 {1174 "epoch": 0.4455179817498658,1175 "grad_norm": 2.3826616991498994,1176 "learning_rate": 6.7871722987505975e-06,1177 "loss": 0.4975,1178 "step": 8301179 },1180 {1181 "epoch": 0.4482018250134192,1182 "grad_norm": 2.350730267510358,1183 "learning_rate": 6.743328733581211e-06,1184 "loss": 0.4974,1185 "step": 8351186 },1187 {1188 "epoch": 0.45088566827697263,1189 "grad_norm": 2.4759950761731533,1190 "learning_rate": 6.6993320357489396e-06,1191 "loss": 0.5109,1192 "step": 8401193 },1194 {1195 "epoch": 0.45356951154052605,1196 "grad_norm": 2.435108148176811,1197 "learning_rate": 6.6551860698900804e-06,1198 "loss": 0.5109,1199 "step": 8451200 },