ANLP-Final-Project/m1-mlp
016
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 1.0,6 "eval_steps": 500,7 "global_step": 100000,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.005,14 "grad_norm": 0.5228317379951477,15 "learning_rate": 3.326666666666667e-06,16 "loss": 0.7898,17 "step": 50018 },19 {20 "epoch": 0.01,21 "grad_norm": 1.0322492122650146,22 "learning_rate": 6.660000000000001e-06,23 "loss": 0.5367,24 "step": 100025 },26 {27 "epoch": 0.015,28 "grad_norm": 2.0602364540100098,29 "learning_rate": 9.993333333333333e-06,30 "loss": 0.4491,31 "step": 150032 },33 {34 "epoch": 0.02,35 "grad_norm": 0.8021494150161743,36 "learning_rate": 1.3326666666666667e-05,37 "loss": 0.3972,38 "step": 200039 },40 {41 "epoch": 0.025,42 "grad_norm": 2.3705506324768066,43 "learning_rate": 1.666e-05,44 "loss": 0.3854,45 "step": 250046 },47 {48 "epoch": 0.03,49 "grad_norm": 1.653393030166626,50 "learning_rate": 1.9993333333333337e-05,51 "loss": 0.3723,52 "step": 300053 },54 {55 "epoch": 0.035,56 "grad_norm": 2.5747907161712646,57 "learning_rate": 1.9998694075951004e-05,58 "loss": 0.3747,59 "step": 350060 },61 {62 "epoch": 0.04,63 "grad_norm": 2.956480026245117,64 "learning_rate": 1.9994766172915367e-05,65 "loss": 0.3435,66 "step": 400067 },68 {69 "epoch": 0.045,70 "grad_norm": 1.4719043970108032,71 "learning_rate": 1.9988217315672814e-05,72 "loss": 0.3456,73 "step": 450074 },75 {76 "epoch": 0.05,77 "grad_norm": 1.938084602355957,78 "learning_rate": 1.997904922154766e-05,79 "loss": 0.3284,80 "step": 500081 },82 {83 "epoch": 0.055,84 "grad_norm": 3.1742770671844482,85 "learning_rate": 1.99672642947137e-05,86 "loss": 0.3323,87 "step": 550088 },89 {90 "epoch": 0.06,91 "grad_norm": 3.088749647140503,92 "learning_rate": 1.995286562556374e-05,93 "loss": 0.333,94 "step": 600095 },96 {97 "epoch": 0.065,98 "grad_norm": 4.315283298492432,99 "learning_rate": 1.993585698989922e-05,100 "loss": 0.335,101 "step": 6500102 },103 {104 "epoch": 0.07,105 "grad_norm": 1.9591349363327026,106 "learning_rate": 1.9916242847940064e-05,107 "loss": 0.3301,108 "step": 7000109 },110 {111 "epoch": 0.075,112 "grad_norm": 2.167367458343506,113 "learning_rate": 1.9894028343155063e-05,114 "loss": 0.3331,115 "step": 7500116 },117 {118 "epoch": 0.08,119 "grad_norm": 1.7126212120056152,120 "learning_rate": 1.9869219300913088e-05,121 "loss": 0.3305,122 "step": 8000123 },124 {125 "epoch": 0.085,126 "grad_norm": 2.834191083908081,127 "learning_rate": 1.9841822226955496e-05,128 "loss": 0.3144,129 "step": 8500130 },131 {132 "epoch": 0.09,133 "grad_norm": 1.9557807445526123,134 "learning_rate": 1.9811844305690107e-05,135 "loss": 0.3165,136 "step": 9000137 },138 {139 "epoch": 0.095,140 "grad_norm": 2.5267953872680664,141 "learning_rate": 1.977929339830722e-05,142 "loss": 0.3174,143 "step": 9500144 },145 {146 "epoch": 0.1,147 "grad_norm": 1.9610834121704102,148 "learning_rate": 1.9744178040718144e-05,149 "loss": 0.3093,150 "step": 10000151 },152 {153 "epoch": 0.105,154 "grad_norm": 3.3462774753570557,155 "learning_rate": 1.9706507441316814e-05,156 "loss": 0.3191,157 "step": 10500158 },159 {160 "epoch": 0.11,161 "grad_norm": 1.884909987449646,162 "learning_rate": 1.9666291478565047e-05,163 "loss": 0.3254,164 "step": 11000165 },166 {167 "epoch": 0.115,168 "grad_norm": 2.515130043029785,169 "learning_rate": 1.962354069840208e-05,170 "loss": 0.3134,171 "step": 11500172 },173 {174 "epoch": 0.12,175 "grad_norm": 3.2403035163879395,176 "learning_rate": 1.9578266311479082e-05,177 "loss": 0.3154,178 "step": 12000179 },180 {181 "epoch": 0.125,182 "grad_norm": 2.7380077838897705,183 "learning_rate": 1.9530480190219365e-05,184 "loss": 0.3203,185 "step": 12500186 },187 {188 "epoch": 0.13,189 "grad_norm": 3.1931233406066895,190 "learning_rate": 1.948019486570503e-05,191 "loss": 0.3058,192 "step": 13000193 },194 {195 "epoch": 0.135,196 "grad_norm": 3.3102500438690186,197 "learning_rate": 1.942742352439093e-05,198 "loss": 0.3024,199 "step": 13500200 },201 {202 "epoch": 0.14,203 "grad_norm": 3.0574374198913574,204 "learning_rate": 1.937218000464673e-05,205 "loss": 0.3115,206 "step": 14000207 },208 {209 "epoch": 0.145,210 "grad_norm": 3.5112788677215576,211 "learning_rate": 1.931447879312804e-05,212 "loss": 0.312,213 "step": 14500214 },215 {216 "epoch": 0.15,217 "grad_norm": 2.5837135314941406,218 "learning_rate": 1.9254335020977542e-05,219 "loss": 0.3095,220 "step": 15000221 },222 {223 "epoch": 0.155,224 "grad_norm": 2.8150858879089355,225 "learning_rate": 1.9191764459857124e-05,226 "loss": 0.3038,227 "step": 15500228 },229 {230 "epoch": 0.16,231 "grad_norm": 2.605210304260254,232 "learning_rate": 1.9126783517812005e-05,233 "loss": 0.2889,234 "step": 16000235 },236 {237 "epoch": 0.165,238 "grad_norm": 2.668278217315674,239 "learning_rate": 1.9059409234968034e-05,240 "loss": 0.2966,241 "step": 16500242 },243 {244 "epoch": 0.17,245 "grad_norm": 1.3171859979629517,246 "learning_rate": 1.8989659279063214e-05,247 "loss": 0.2891,248 "step": 17000249 },250 {251 "epoch": 0.175,252 "grad_norm": 3.820676565170288,253 "learning_rate": 1.8917551940814615e-05,254 "loss": 0.2953,255 "step": 17500256 },257 {258 "epoch": 0.18,259 "grad_norm": 2.1207947731018066,260 "learning_rate": 1.8843106129121993e-05,261 "loss": 0.2982,262 "step": 18000263 },264 {265 "epoch": 0.185,266 "grad_norm": 19.94713592529297,267 "learning_rate": 1.8766341366109214e-05,268 "loss": 0.2908,269 "step": 18500270 },271 {272 "epoch": 0.19,273 "grad_norm": 2.073420286178589,274 "learning_rate": 1.8687277782004954e-05,275 "loss": 0.2788,276 "step": 19000277 },278 {279 "epoch": 0.195,280 "grad_norm": 3.1489017009735107,281 "learning_rate": 1.8605936109863872e-05,282 "loss": 0.299,283 "step": 19500284 },285 {286 "epoch": 0.2,287 "grad_norm": 2.1240925788879395,288 "learning_rate": 1.8522337680129728e-05,289 "loss": 0.2905,290 "step": 20000291 },292 {293 "epoch": 0.205,294 "grad_norm": 2.2307348251342773,295 "learning_rate": 1.843650441504186e-05,296 "loss": 0.2928,297 "step": 20500298 },299 {300 "epoch": 0.21,301 "grad_norm": 1.1732121706008911,302 "learning_rate": 1.8348458822886437e-05,303 "loss": 0.2893,304 "step": 21000305 },306 {307 "epoch": 0.215,308 "grad_norm": 2.374523639678955,309 "learning_rate": 1.8258223992094055e-05,310 "loss": 0.2995,311 "step": 21500312 },313 {314 "epoch": 0.22,315 "grad_norm": 1.002447247505188,316 "learning_rate": 1.816582358518521e-05,317 "loss": 0.2884,318 "step": 22000319 },320 {321 "epoch": 0.225,322 "grad_norm": 2.586608409881592,323 "learning_rate": 1.80712818325652e-05,324 "loss": 0.2908,325 "step": 22500326 },327 {328 "epoch": 0.23,329 "grad_norm": 2.567784070968628,330 "learning_rate": 1.7974623526170113e-05,331 "loss": 0.2955,332 "step": 23000333 },334 {335 "epoch": 0.235,336 "grad_norm": 1.543729305267334,337 "learning_rate": 1.78758740129656e-05,338 "loss": 0.2848,339 "step": 23500340 },341 {342 "epoch": 0.24,343 "grad_norm": 4.3870954513549805,344 "learning_rate": 1.777505918830004e-05,345 "loss": 0.2856,346 "step": 24000347 },348 {349 "epoch": 0.245,350 "grad_norm": 2.431323528289795,351 "learning_rate": 1.767220548911395e-05,352 "loss": 0.2699,353 "step": 24500354 },355 {356 "epoch": 0.25,357 "grad_norm": 2.8813323974609375,358 "learning_rate": 1.756733988700737e-05,359 "loss": 0.275,360 "step": 25000361 },362 {363 "epoch": 0.255,364 "grad_norm": 2.1704187393188477,365 "learning_rate": 1.7460489881166986e-05,366 "loss": 0.2852,367 "step": 25500368 },369 {370 "epoch": 0.26,371 "grad_norm": 1.9194124937057495,372 "learning_rate": 1.7351683491154984e-05,373 "loss": 0.2963,374 "step": 26000375 },376 {377 "epoch": 0.265,378 "grad_norm": 2.9418258666992188,379 "learning_rate": 1.7240949249561402e-05,380 "loss": 0.282,381 "step": 26500382 },383 {384 "epoch": 0.27,385 "grad_norm": 2.2254602909088135,386 "learning_rate": 1.7128316194521923e-05,387 "loss": 0.269,388 "step": 27000389 },390 {391 "epoch": 0.275,392 "grad_norm": 3.698273181915283,393 "learning_rate": 1.7013813862103163e-05,394 "loss": 0.2748,395 "step": 27500396 },397 {398 "epoch": 0.28,399 "grad_norm": 1.1656277179718018,400 "learning_rate": 1.689747227855733e-05,401 "loss": 0.2798,402 "step": 28000403 },404 {405 "epoch": 0.285,406 "grad_norm": 2.6570346355438232,407 "learning_rate": 1.677932195244836e-05,408 "loss": 0.2791,409 "step": 28500410 },411 {412 "epoch": 0.29,413 "grad_norm": 3.1429874897003174,414 "learning_rate": 1.6659393866651585e-05,415 "loss": 0.2695,416 "step": 29000417 },418 {419 "epoch": 0.295,420 "grad_norm": 3.390726089477539,421 "learning_rate": 1.6537719470228993e-05,422 "loss": 0.2749,423 "step": 29500424 },425 {426 "epoch": 0.3,427 "grad_norm": 2.2655887603759766,428 "learning_rate": 1.6414330670182278e-05,429 "loss": 0.2764,430 "step": 30000431 },432 {433 "epoch": 0.305,434 "grad_norm": 2.555089235305786,435 "learning_rate": 1.6289259823085765e-05,436 "loss": 0.2834,437 "step": 30500438 },439 {440 "epoch": 0.31,441 "grad_norm": 2.014594793319702,442 "learning_rate": 1.6162539726601477e-05,443 "loss": 0.2801,444 "step": 31000445 },446 {447 "epoch": 0.315,448 "grad_norm": 3.4530141353607178,449 "learning_rate": 1.603420361087849e-05,450 "loss": 0.2789,451 "step": 31500452 },453 {454 "epoch": 0.32,455 "grad_norm": 1.4499799013137817,456 "learning_rate": 1.5904285129838924e-05,457 "loss": 0.2799,458 "step": 32000459 },460 {461 "epoch": 0.325,462 "grad_norm": 2.807821035385132,463 "learning_rate": 1.577281835235279e-05,464 "loss": 0.2717,465 "step": 32500466 },467 {468 "epoch": 0.33,469 "grad_norm": 2.2752676010131836,470 "learning_rate": 1.563983775330399e-05,471 "loss": 0.2694,472 "step": 33000473 },474 {475 "epoch": 0.335,476 "grad_norm": 1.8416115045547485,477 "learning_rate": 1.5505378204549907e-05,478 "loss": 0.2773,479 "step": 33500480 },481 {482 "epoch": 0.34,483 "grad_norm": 5.743653297424316,484 "learning_rate": 1.536947496577684e-05,485 "loss": 0.2786,486 "step": 34000487 },488 {489 "epoch": 0.345,490 "grad_norm": 2.5975189208984375,491 "learning_rate": 1.5232163675253793e-05,492 "loss": 0.2825,493 "step": 34500494 },495 {496 "epoch": 0.35,497 "grad_norm": 1.8504586219787598,498 "learning_rate": 1.5093480340486922e-05,499 "loss": 0.2693,500 "step": 35000501 },502 {503 "epoch": 0.355,504 "grad_norm": 2.5584330558776855,505 "learning_rate": 1.495346132877721e-05,506 "loss": 0.2688,507 "step": 35500508 },509 {510 "epoch": 0.36,511 "grad_norm": 3.1796674728393555,512 "learning_rate": 1.4812143357683773e-05,513 "loss": 0.2552,514 "step": 36000515 },516 {517 "epoch": 0.365,518 "grad_norm": 2.1949145793914795,519 "learning_rate": 1.4669563485395315e-05,520 "loss": 0.2735,521 "step": 36500522 },523 {524 "epoch": 0.37,525 "grad_norm": 2.1829614639282227,526 "learning_rate": 1.4525759101012249e-05,527 "loss": 0.2778,528 "step": 37000529 },530 {531 "epoch": 0.375,532 "grad_norm": 2.5891077518463135,533 "learning_rate": 1.4380767914742074e-05,534 "loss": 0.2703,535 "step": 37500536 },537 {538 "epoch": 0.38,539 "grad_norm": 1.912104606628418,540 "learning_rate": 1.4234627948010497e-05,541 "loss": 0.2603,542 "step": 38000543 },544 {545 "epoch": 0.385,546 "grad_norm": 2.5725393295288086,547 "learning_rate": 1.4087377523491016e-05,548 "loss": 0.2745,549 "step": 38500550 },551 {552 "epoch": 0.39,553 "grad_norm": 2.139127016067505,554 "learning_rate": 1.3939055255055419e-05,555 "loss": 0.2592,556 "step": 39000557 },558 {559 "epoch": 0.395,560 "grad_norm": 3.271855115890503,561 "learning_rate": 1.3789700037647988e-05,562 "loss": 0.2676,563 "step": 39500564 },565 {566 "epoch": 0.4,567 "grad_norm": 4.065393924713135,568 "learning_rate": 1.363935103708598e-05,569 "loss": 0.2582,570 "step": 40000571 },572 {573 "epoch": 0.405,574 "grad_norm": 4.319698333740234,575 "learning_rate": 1.3488047679789052e-05,576 "loss": 0.2738,577 "step": 40500578 },579 {580 "epoch": 0.41,581 "grad_norm": 1.9484474658966064,582 "learning_rate": 1.3335829642440372e-05,583 "loss": 0.2625,584 "step": 41000585 },586 {587 "epoch": 0.415,588 "grad_norm": 2.7909133434295654,589 "learning_rate": 1.3182736841582109e-05,590 "loss": 0.2732,591 "step": 41500592 },593 {594 "epoch": 0.42,595 "grad_norm": 1.9949915409088135,596 "learning_rate": 1.3028809423147997e-05,597 "loss": 0.2789,598 "step": 42000599 },600 {601 "epoch": 0.425,602 "grad_norm": 2.6040537357330322,603 "learning_rate": 1.2874087751935756e-05,604 "loss": 0.27,605 "step": 42500606 },607 {608 "epoch": 0.43,609 "grad_norm": 2.1791725158691406,610 "learning_rate": 1.2718612401022148e-05,611 "loss": 0.2682,612 "step": 43000613 },614 {615 "epoch": 0.435,616 "grad_norm": 2.334940195083618,617 "learning_rate": 1.2562424141123364e-05,618 "loss": 0.2702,619 "step": 43500620 },621 {622 "epoch": 0.44,623 "grad_norm": 3.011237144470215,624 "learning_rate": 1.2405563929903614e-05,625 "loss": 0.2684,626 "step": 44000627 },628 {629 "epoch": 0.445,630 "grad_norm": 3.324738025665283,631 "learning_rate": 1.2248072901234715e-05,632 "loss": 0.2626,633 "step": 44500634 },635 {636 "epoch": 0.45,637 "grad_norm": 2.0907318592071533,638 "learning_rate": 1.2089992354409415e-05,639 "loss": 0.2673,640 "step": 45000641 },642 {643 "epoch": 0.455,644 "grad_norm": 2.5054478645324707,645 "learning_rate": 1.1931363743311377e-05,646 "loss": 0.2634,647 "step": 45500648 },649 {650 "epoch": 0.46,651 "grad_norm": 3.015181541442871,652 "learning_rate": 1.1772228665544634e-05,653 "loss": 0.2623,654 "step": 46000655 },656 {657 "epoch": 0.465,658 "grad_norm": 1.9461313486099243,659 "learning_rate": 1.1612628851525313e-05,660 "loss": 0.2685,661 "step": 46500662 },663 {664 "epoch": 0.47,665 "grad_norm": 2.5035080909729004,666 "learning_rate": 1.1452606153538577e-05,667 "loss": 0.2625,668 "step": 47000669 },670 {671 "epoch": 0.475,672 "grad_norm": 1.4080424308776855,673 "learning_rate": 1.129220253476358e-05,674 "loss": 0.2651,675 "step": 47500676 },677 {678 "epoch": 0.48,679 "grad_norm": 2.9700748920440674,680 "learning_rate": 1.1131460058269342e-05,681 "loss": 0.2547,682 "step": 48000683 },684 {685 "epoch": 0.485,686 "grad_norm": 2.252192258834839,687 "learning_rate": 1.0970420875984453e-05,688 "loss": 0.2629,689 "step": 48500690 },691 {692 "epoch": 0.49,693 "grad_norm": 3.6359364986419678,694 "learning_rate": 1.0809127217643462e-05,695 "loss": 0.2623,696 "step": 49000697 },698 {699 "epoch": 0.495,700 "grad_norm": 2.2924039363861084,701 "learning_rate": 1.0647621379712843e-05,702 "loss": 0.2648,703 "step": 49500704 },705 {706 "epoch": 0.5,707 "grad_norm": 1.8426686525344849,708 "learning_rate": 1.0485945714299487e-05,709 "loss": 0.2536,710 "step": 50000711 },712 {713 "epoch": 0.505,714 "grad_norm": 2.85508131980896,715 "learning_rate": 1.0324142618044612e-05,716 "loss": 0.2685,717 "step": 50500718 },719 {720 "epoch": 0.51,721 "grad_norm": 1.8583101034164429,722 "learning_rate": 1.0162254521005956e-05,723 "loss": 0.2627,724 "step": 51000725 },726 {727 "epoch": 0.515,728 "grad_norm": 2.0837361812591553,729 "learning_rate": 1.0000323875531241e-05,730 "loss": 0.2594,731 "step": 51500732 },733 {734 "epoch": 0.52,735 "grad_norm": 2.786468505859375,736 "learning_rate": 9.838393145125787e-06,737 "loss": 0.2602,738 "step": 52000739 },740 {741 "epoch": 0.525,742 "grad_norm": 3.2767674922943115,743 "learning_rate": 9.676504793317163e-06,744 "loss": 0.2575,745 "step": 52500746 },747 {748 "epoch": 0.53,749 "grad_norm": 2.087574005126953,750 "learning_rate": 9.514701272519918e-06,751 "loss": 0.2536,752 "step": 53000753 },754 {755 "epoch": 0.535,756 "grad_norm": 1.6230825185775757,757 "learning_rate": 9.353025012903112e-06,758 "loss": 0.2634,759 "step": 53500760 },761 {762 "epoch": 0.54,763 "grad_norm": 1.9455654621124268,764 "learning_rate": 9.191518411263777e-06,765 "loss": 0.2415,766 "step": 54000767 },768 {769 "epoch": 0.545,770 "grad_norm": 2.6995887756347656,771 "learning_rate": 9.03022381990908e-06,772 "loss": 0.2687,773 "step": 54500774 },775 {776 "epoch": 0.55,777 "grad_norm": 4.489888668060303,778 "learning_rate": 8.869183535550167e-06,779 "loss": 0.2719,780 "step": 55000781 },782 {783 "epoch": 0.555,784 "grad_norm": 2.966480016708374,785 "learning_rate": 8.708439788210573e-06,786 "loss": 0.268,787 "step": 55500788 },789 {790 "epoch": 0.56,791 "grad_norm": 3.2469143867492676,792 "learning_rate": 8.548034730152153e-06,793 "loss": 0.262,794 "step": 56000795 },796 {797 "epoch": 0.565,798 "grad_norm": 2.2795863151550293,799 "learning_rate": 8.388010424821354e-06,800 "loss": 0.2644,801 "step": 56500802 },803 {804 "epoch": 0.57,805 "grad_norm": 2.264314889907837,806 "learning_rate": 8.22840883581883e-06,807 "loss": 0.2519,808 "step": 57000809 },810 {811 "epoch": 0.575,812 "grad_norm": 2.3506600856781006,813 "learning_rate": 8.069271815895196e-06,814 "loss": 0.2797,815 "step": 57500816 },817 {818 "epoch": 0.58,819 "grad_norm": 4.5607428550720215,820 "learning_rate": 7.910641095975886e-06,821 "loss": 0.2574,822 "step": 58000823 },824 {825 "epoch": 0.585,826 "grad_norm": 0.9732487201690674,827 "learning_rate": 7.752558274217919e-06,828 "loss": 0.2529,829 "step": 58500830 },831 {832 "epoch": 0.59,833 "grad_norm": 3.291529417037964,834 "learning_rate": 7.595064805101538e-06,835 "loss": 0.2502,836 "step": 59000837 },838 {839 "epoch": 0.595,840 "grad_norm": 1.8199131488800049,841 "learning_rate": 7.438201988559449e-06,842 "loss": 0.2643,843 "step": 59500844 },845 {846 "epoch": 0.6,847 "grad_norm": 5.467274188995361,848 "learning_rate": 7.282010959146662e-06,849 "loss": 0.2636,850 "step": 60000851 },852 {853 "epoch": 0.605,854 "grad_norm": 3.899115562438965,855 "learning_rate": 7.1265326752536545e-06,856 "loss": 0.2625,857 "step": 60500858 },859 {860 "epoch": 0.61,861 "grad_norm": 2.5981805324554443,862 "learning_rate": 6.971807908365737e-06,863 "loss": 0.2577,864 "step": 61000865 },866 {867 "epoch": 0.615,868 "grad_norm": 2.875891923904419,869 "learning_rate": 6.817877232371453e-06,870 "loss": 0.2649,871 "step": 61500872 },873 {874 "epoch": 0.62,875 "grad_norm": 2.9105169773101807,876 "learning_rate": 6.66478101292277e-06,877 "loss": 0.2496,878 "step": 62000879 },880 {881 "epoch": 0.625,882 "grad_norm": 3.293264389038086,883 "learning_rate": 6.5125593968498855e-06,884 "loss": 0.2444,885 "step": 62500886 },887 {888 "epoch": 0.63,889 "grad_norm": 3.560056209564209,890 "learning_rate": 6.361252301633422e-06,891 "loss": 0.2625,892 "step": 63000893 },894 {895 "epoch": 0.635,896 "grad_norm": 2.613802909851074,897 "learning_rate": 6.210899404936761e-06,898 "loss": 0.2526,899 "step": 63500900 },901 {902 "epoch": 0.64,903 "grad_norm": 3.4607479572296143,904 "learning_rate": 6.061540134201248e-06,905 "loss": 0.2556,906 "step": 64000907 },908 {909 "epoch": 0.645,910 "grad_norm": 3.917365074157715,911 "learning_rate": 5.913213656307049e-06,912 "loss": 0.2601,913 "step": 64500914 },915 {916 "epoch": 0.65,917 "grad_norm": 2.4207093715667725,918 "learning_rate": 5.765958867302311e-06,919 "loss": 0.2512,920 "step": 65000921 },922 {923 "epoch": 0.655,924 "grad_norm": 2.076265335083008,925 "learning_rate": 5.619814382203341e-06,926 "loss": 0.2437,927 "step": 65500928 },929 {930 "epoch": 0.66,931 "grad_norm": 3.774073600769043,932 "learning_rate": 5.474818524868482e-06,933 "loss": 0.2726,934 "step": 66000935 },936 {937 "epoch": 0.665,938 "grad_norm": 3.220529079437256,939 "learning_rate": 5.331009317948359e-06,940 "loss": 0.2664,941 "step": 66500942 },943 {944 "epoch": 0.67,945 "grad_norm": 2.737941026687622,946 "learning_rate": 5.188424472915091e-06,947 "loss": 0.2703,948 "step": 67000949 },950 {951 "epoch": 0.675,952 "grad_norm": 2.459578275680542,953 "learning_rate": 5.047101380173111e-06,954 "loss": 0.244,955 "step": 67500956 },957 {958 "epoch": 0.68,959 "grad_norm": 3.3251523971557617,960 "learning_rate": 4.907077099254194e-06,961 "loss": 0.2597,962 "step": 68000963 },964 {965 "epoch": 0.685,966 "grad_norm": 2.779625654220581,967 "learning_rate": 4.7683883490991985e-06,968 "loss": 0.2623,969 "step": 68500970 },971 {972 "epoch": 0.69,973 "grad_norm": 1.0731335878372192,974 "learning_rate": 4.6310714984292096e-06,975 "loss": 0.2616,976 "step": 69000977 },978 {979 "epoch": 0.695,980 "grad_norm": 1.1064262390136719,981 "learning_rate": 4.495162556208438e-06,982 "loss": 0.2489,983 "step": 69500984 },985 {986 "epoch": 0.7,987 "grad_norm": 6.329731464385986,988 "learning_rate": 4.360697162201508e-06,989 "loss": 0.254,990 "step": 70000991 },992 {993 "epoch": 0.705,994 "grad_norm": 3.1738226413726807,995 "learning_rate": 4.2277105776275395e-06,996 "loss": 0.249,997 "step": 70500998 },999 {1000 "epoch": 0.71,1001 "grad_norm": 2.834291696548462,1002 "learning_rate": 4.096237675913521e-06,1003 "loss": 0.2607,1004 "step": 710001005 },1006 {1007 "epoch": 0.715,1008 "grad_norm": 2.2489569187164307,1009 "learning_rate": 3.966312933549325e-06,1010 "loss": 0.24,1011 "step": 715001012 },1013 {1014 "epoch": 0.72,1015 "grad_norm": 4.554741859436035,1016 "learning_rate": 3.837970421046881e-06,1017 "loss": 0.2492,1018 "step": 720001019 },1020 {1021 "epoch": 0.725,1022 "grad_norm": 4.065186977386475,1023 "learning_rate": 3.7112437940057543e-06,1024 "loss": 0.2491,1025 "step": 725001026 },1027 {1028 "epoch": 0.73,1029 "grad_norm": 3.260732650756836,1030 "learning_rate": 3.5861662842875534e-06,1031 "loss": 0.2413,1032 "step": 730001033 },1034 {1035 "epoch": 0.735,1036 "grad_norm": 4.497494220733643,1037 "learning_rate": 3.4627706913014503e-06,1038 "loss": 0.2537,1039 "step": 735001040 },1041 {1042 "epoch": 0.74,1043 "grad_norm": 2.5874199867248535,1044 "learning_rate": 3.3410893734031046e-06,1045 "loss": 0.2627,1046 "step": 740001047 },1048 {1049 "epoch": 0.745,1050 "grad_norm": 2.394810438156128,1051 "learning_rate": 3.2211542394092443e-06,1052 "loss": 0.2516,1053 "step": 745001054 },1055 {1056 "epoch": 0.75,1057 "grad_norm": 2.5389678478240967,1058 "learning_rate": 3.102996740230131e-06,1059 "loss": 0.2434,1060 "step": 750001061 },1062 {1063 "epoch": 0.755,1064 "grad_norm": 1.8920937776565552,1065 "learning_rate": 2.986647860622105e-06,1066 "loss": 0.2602,1067 "step": 755001068 },1069 {1070 "epoch": 0.76,1071 "grad_norm": 1.569104790687561,1072 "learning_rate": 2.8721381110623724e-06,1073 "loss": 0.248,1074 "step": 760001075 },1076 {1077 "epoch": 0.765,1078 "grad_norm": 2.620171308517456,1079 "learning_rate": 2.7594975197481546e-06,1080 "loss": 0.249,1081 "step": 765001082 },1083 {1084 "epoch": 0.77,1085 "grad_norm": 3.781393527984619,1086 "learning_rate": 2.64875562472232e-06,1087 "loss": 0.2475,1088 "step": 770001089 },1090 {1091 "epoch": 0.775,1092 "grad_norm": 4.293469429016113,1093 "learning_rate": 2.5399414661275444e-06,1094 "loss": 0.2584,1095 "step": 775001096 },1097 {1098 "epoch": 0.78,1099 "grad_norm": 1.6750516891479492,1100 "learning_rate": 2.433083578591029e-06,1101 "loss": 0.2471,1102 "step": 780001103 },1104 {1105 "epoch": 0.785,1106 "grad_norm": 3.109180450439453,1107 "learning_rate": 2.3282099837418004e-06,1108 "loss": 0.257,1109 "step": 785001110 },1111 {1112 "epoch": 0.79,1113 "grad_norm": 3.492893695831299,1114 "learning_rate": 2.2253481828625144e-06,1115 "loss": 0.2644,1116 "step": 790001117 },1118 {1119 "epoch": 0.795,1120 "grad_norm": 3.8731396198272705,1121 "learning_rate": 2.1245251496777274e-06,1122 "loss": 0.2546,1123 "step": 795001124 },1125 {1126 "epoch": 0.8,1127 "grad_norm": 2.1617014408111572,1128 "learning_rate": 2.025767323280502e-06,1129 "loss": 0.2542,1130 "step": 800001131 },1132 {1133 "epoch": 0.805,1134 "grad_norm": 5.593191146850586,1135 "learning_rate": 1.929100601199211e-06,1136 "loss": 0.2376,1137 "step": 805001138 },1139 {1140 "epoch": 0.81,1141 "grad_norm": 3.496340036392212,1142 "learning_rate": 1.834550332606365e-06,1143 "loss": 0.234,1144 "step": 810001145 },1146 {1147 "epoch": 0.815,1148 "grad_norm": 2.7082982063293457,1149 "learning_rate": 1.742141311671227e-06,1150 "loss": 0.2482,1151 "step": 815001152 },1153 {1154 "epoch": 0.82,1155 "grad_norm": 2.0201199054718018,1156 "learning_rate": 1.6518977710579708e-06,1157 "loss": 0.2676,1158 "step": 820001159 },1160 {1161 "epoch": 0.825,1162 "grad_norm": 2.157087564468384,1163 "learning_rate": 1.5638433755710892e-06,1164 "loss": 0.2505,1165 "step": 825001166 },1167 {1168 "epoch": 0.83,1169 "grad_norm": 2.467223882675171,1170 "learning_rate": 1.4780012159497037e-06,1171 "loss": 0.2632,1172 "step": 830001173 },1174 {1175 "epoch": 0.835,1176 "grad_norm": 1.9714136123657227,1177 "learning_rate": 1.3943938028124216e-06,1178 "loss": 0.2568,1179 "step": 835001180 },1181 {1182 "epoch": 0.84,1183 "grad_norm": 6.337538242340088,1184 "learning_rate": 1.3130430607543166e-06,1185 "loss": 0.2555,1186 "step": 840001187 },1188 {1189 "epoch": 0.845,1190 "grad_norm": 5.382758617401123,1191 "learning_rate": 1.2339703225975818e-06,1192 "loss": 0.2508,1193 "step": 845001194 },1195 {1196 "epoch": 0.85,1197 "grad_norm": 2.141634464263916,1198 "learning_rate": 1.1571963237973693e-06,1199 "loss": 0.246,1200 "step": 85000