lldois/v12_user_cot_focus_lr15e6
08
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": 927,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.005398110661268556,14 "grad_norm": 46.5,15 "learning_rate": 2.1428571428571427e-06,16 "loss": 2.6865272521972656,17 "step": 518 },19 {20 "epoch": 0.010796221322537112,21 "grad_norm": 25.5,22 "learning_rate": 4.821428571428572e-06,23 "loss": 2.7759025573730467,24 "step": 1025 },26 {27 "epoch": 0.016194331983805668,28 "grad_norm": 9.625,29 "learning_rate": 7.5e-06,30 "loss": 2.3364213943481444,31 "step": 1532 },33 {34 "epoch": 0.021592442645074223,35 "grad_norm": 10.0625,36 "learning_rate": 1.0178571428571429e-05,37 "loss": 2.357875442504883,38 "step": 2039 },40 {41 "epoch": 0.02699055330634278,42 "grad_norm": 7.125,43 "learning_rate": 1.2857142857142857e-05,44 "loss": 2.113478088378906,45 "step": 2546 },47 {48 "epoch": 0.032388663967611336,49 "grad_norm": 5.34375,50 "learning_rate": 1.4999954205725012e-05,51 "loss": 2.0126585006713866,52 "step": 3053 },54 {55 "epoch": 0.037786774628879895,56 "grad_norm": 3.546875,57 "learning_rate": 1.4998351464818896e-05,58 "loss": 1.952888298034668,59 "step": 3560 },61 {62 "epoch": 0.043184885290148446,63 "grad_norm": 3.875,64 "learning_rate": 1.4994459569363502e-05,65 "loss": 1.7912431716918946,66 "step": 4067 },68 {69 "epoch": 0.048582995951417005,70 "grad_norm": 4.5625,71 "learning_rate": 1.4988279707506684e-05,72 "loss": 1.7172063827514648,73 "step": 4574 },75 {76 "epoch": 0.05398110661268556,77 "grad_norm": 3.828125,78 "learning_rate": 1.4979813765884318e-05,79 "loss": 1.6440818786621094,80 "step": 5081 },82 {83 "epoch": 0.059379217273954114,84 "grad_norm": 5.28125,85 "learning_rate": 1.4969064329044335e-05,86 "loss": 1.691044235229492,87 "step": 5588 },89 {90 "epoch": 0.06477732793522267,91 "grad_norm": 3.1875,92 "learning_rate": 1.4956034678657697e-05,93 "loss": 1.6486940383911133,94 "step": 6095 },96 {97 "epoch": 0.07017543859649122,98 "grad_norm": 4.3125,99 "learning_rate": 1.4940728792516532e-05,100 "loss": 1.621745491027832,101 "step": 65102 },103 {104 "epoch": 0.07557354925775979,105 "grad_norm": 3.140625,106 "learning_rate": 1.492315134331978e-05,107 "loss": 1.7037588119506837,108 "step": 70109 },110 {111 "epoch": 0.08097165991902834,112 "grad_norm": 4.46875,113 "learning_rate": 1.490330769724665e-05,114 "loss": 1.5873713493347168,115 "step": 75116 },117 {118 "epoch": 0.08636977058029689,119 "grad_norm": 3.078125,120 "learning_rate": 1.4881203912318417e-05,121 "loss": 1.5764989852905273,122 "step": 80123 },124 {125 "epoch": 0.09176788124156546,126 "grad_norm": 2.515625,127 "learning_rate": 1.4856846736548972e-05,128 "loss": 1.6242074966430664,129 "step": 85130 },131 {132 "epoch": 0.09716599190283401,133 "grad_norm": 2.5625,134 "learning_rate": 1.4830243605884732e-05,135 "loss": 1.5501031875610352,136 "step": 90137 },138 {139 "epoch": 0.10256410256410256,140 "grad_norm": 2.46875,141 "learning_rate": 1.4801402641934543e-05,142 "loss": 1.7399942398071289,143 "step": 95144 },145 {146 "epoch": 0.10796221322537113,147 "grad_norm": 3.46875,148 "learning_rate": 1.4770332649490248e-05,149 "loss": 1.5935526847839356,150 "step": 100151 },152 {153 "epoch": 0.11336032388663968,154 "grad_norm": 2.453125,155 "learning_rate": 1.4737043113838703e-05,156 "loss": 1.6135421752929688,157 "step": 105158 },159 {160 "epoch": 0.11875843454790823,161 "grad_norm": 3.734375,162 "learning_rate": 1.4701544197866017e-05,163 "loss": 1.4750379562377929,164 "step": 110165 },166 {167 "epoch": 0.1241565452091768,168 "grad_norm": 3.421875,169 "learning_rate": 1.4663846738954966e-05,170 "loss": 1.5500977516174317,171 "step": 115172 },173 {174 "epoch": 0.12955465587044535,175 "grad_norm": 3.15625,176 "learning_rate": 1.4623962245676462e-05,177 "loss": 1.4977538108825683,178 "step": 120179 },180 {181 "epoch": 0.1349527665317139,182 "grad_norm": 2.140625,183 "learning_rate": 1.4581902894276133e-05,184 "loss": 1.5538610458374023,185 "step": 125186 },187 {188 "epoch": 0.14035087719298245,189 "grad_norm": 3.515625,190 "learning_rate": 1.4537681524957061e-05,191 "loss": 1.5750856399536133,192 "step": 130193 },194 {195 "epoch": 0.145748987854251,196 "grad_norm": 2.71875,197 "learning_rate": 1.4491311637959828e-05,198 "loss": 1.5172950744628906,199 "step": 135200 },201 {202 "epoch": 0.15114709851551958,203 "grad_norm": 2.65625,204 "learning_rate": 1.4442807389441062e-05,205 "loss": 1.4594326972961427,206 "step": 140207 },208 {209 "epoch": 0.15654520917678813,210 "grad_norm": 2.921875,211 "learning_rate": 1.439218358715173e-05,212 "loss": 1.5347204208374023,213 "step": 145214 },215 {216 "epoch": 0.16194331983805668,217 "grad_norm": 5.0625,218 "learning_rate": 1.4339455685916516e-05,219 "loss": 1.490595531463623,220 "step": 150221 },222 {223 "epoch": 0.16734143049932523,224 "grad_norm": 3.421875,225 "learning_rate": 1.4284639782915652e-05,226 "loss": 1.4158689498901367,227 "step": 155228 },229 {230 "epoch": 0.17273954116059378,231 "grad_norm": 2.8125,232 "learning_rate": 1.4227752612770654e-05,233 "loss": 1.4857367515563964,234 "step": 160235 },236 {237 "epoch": 0.17813765182186234,238 "grad_norm": 2.96875,239 "learning_rate": 1.4168811542435435e-05,240 "loss": 1.5408924102783204,241 "step": 165242 },243 {244 "epoch": 0.18353576248313092,245 "grad_norm": 2.90625,246 "learning_rate": 1.4107834565894396e-05,247 "loss": 1.6404100418090821,248 "step": 170249 },250 {251 "epoch": 0.18893387314439947,252 "grad_norm": 2.828125,253 "learning_rate": 1.4044840298669085e-05,254 "loss": 1.5440165519714355,255 "step": 175256 },257 {258 "epoch": 0.19433198380566802,259 "grad_norm": 4.03125,260 "learning_rate": 1.3979847972135099e-05,261 "loss": 1.4469735145568847,262 "step": 180263 },264 {265 "epoch": 0.19973009446693657,266 "grad_norm": 2.65625,267 "learning_rate": 1.391287742765099e-05,268 "loss": 1.4575271606445312,269 "step": 185270 },271 {272 "epoch": 0.20512820512820512,273 "grad_norm": 2.96875,274 "learning_rate": 1.3843949110500939e-05,275 "loss": 1.468193817138672,276 "step": 190277 },278 {279 "epoch": 0.21052631578947367,280 "grad_norm": 2.859375,281 "learning_rate": 1.3773084063653057e-05,282 "loss": 1.380112075805664,283 "step": 195284 },285 {286 "epoch": 0.21592442645074225,287 "grad_norm": 3.71875,288 "learning_rate": 1.3700303921335235e-05,289 "loss": 1.4284241676330567,290 "step": 200291 },292 {293 "epoch": 0.2213225371120108,294 "grad_norm": 2.828125,295 "learning_rate": 1.3625630902430466e-05,296 "loss": 1.4978090286254884,297 "step": 205298 },299 {300 "epoch": 0.22672064777327935,301 "grad_norm": 3.3125,302 "learning_rate": 1.3549087803693715e-05,303 "loss": 1.4717984199523926,304 "step": 210305 },306 {307 "epoch": 0.2321187584345479,308 "grad_norm": 3.203125,309 "learning_rate": 1.3470697992792328e-05,310 "loss": 1.5032032966613769,311 "step": 215312 },313 {314 "epoch": 0.23751686909581646,315 "grad_norm": 2.578125,316 "learning_rate": 1.339048540117218e-05,317 "loss": 1.539174461364746,318 "step": 220319 },320 {321 "epoch": 0.242914979757085,322 "grad_norm": 2.703125,323 "learning_rate": 1.3308474516751701e-05,324 "loss": 1.4706098556518554,325 "step": 225326 },327 {328 "epoch": 0.2483130904183536,329 "grad_norm": 3.1875,330 "learning_rate": 1.322469037644601e-05,331 "loss": 1.4873103141784667,332 "step": 230333 },334 {335 "epoch": 0.25371120107962214,336 "grad_norm": 3.046875,337 "learning_rate": 1.3139158558523457e-05,338 "loss": 1.4533129692077638,339 "step": 235340 },341 {342 "epoch": 0.2591093117408907,343 "grad_norm": 2.640625,344 "learning_rate": 1.3051905174796882e-05,345 "loss": 1.5055602073669434,346 "step": 240347 },348 {349 "epoch": 0.26450742240215924,350 "grad_norm": 2.375,351 "learning_rate": 1.296295686265202e-05,352 "loss": 1.4806604385375977,353 "step": 245354 },355 {356 "epoch": 0.2699055330634278,357 "grad_norm": 5.21875,358 "learning_rate": 1.2872340776915414e-05,359 "loss": 1.463977336883545,360 "step": 250361 },362 {363 "epoch": 0.27530364372469635,364 "grad_norm": 2.515625,365 "learning_rate": 1.2780084581564391e-05,366 "loss": 1.445403480529785,367 "step": 255368 },369 {370 "epoch": 0.2807017543859649,371 "grad_norm": 3.015625,372 "learning_rate": 1.2686216441281581e-05,373 "loss": 1.4426302909851074,374 "step": 260375 },376 {377 "epoch": 0.28609986504723345,378 "grad_norm": 2.9375,379 "learning_rate": 1.2590765012856599e-05,380 "loss": 1.4424419403076172,381 "step": 265382 },383 {384 "epoch": 0.291497975708502,385 "grad_norm": 2.546875,386 "learning_rate": 1.2493759436437467e-05,387 "loss": 1.3920962333679199,388 "step": 270389 },390 {391 "epoch": 0.2968960863697706,392 "grad_norm": 4.0625,393 "learning_rate": 1.2395229326634489e-05,394 "loss": 1.481032943725586,395 "step": 275396 },397 {398 "epoch": 0.30229419703103916,399 "grad_norm": 3.171875,400 "learning_rate": 1.2295204763479281e-05,401 "loss": 1.3904236793518066,402 "step": 280403 },404 {405 "epoch": 0.3076923076923077,406 "grad_norm": 3.34375,407 "learning_rate": 1.2193716283241692e-05,408 "loss": 1.4106941223144531,409 "step": 285410 },411 {412 "epoch": 0.31309041835357626,413 "grad_norm": 3.015625,414 "learning_rate": 1.2090794869107473e-05,415 "loss": 1.4160966873168945,416 "step": 290417 },418 {419 "epoch": 0.3184885290148448,420 "grad_norm": 4.375,421 "learning_rate": 1.1986471941719478e-05,422 "loss": 1.392179298400879,423 "step": 295424 },425 {426 "epoch": 0.32388663967611336,427 "grad_norm": 3.03125,428 "learning_rate": 1.1880779349585331e-05,429 "loss": 1.4000489234924316,430 "step": 300431 },432 {433 "epoch": 0.3292847503373819,434 "grad_norm": 2.40625,435 "learning_rate": 1.1773749359354481e-05,436 "loss": 1.482375717163086,437 "step": 305438 },439 {440 "epoch": 0.33468286099865047,441 "grad_norm": 2.140625,442 "learning_rate": 1.166541464596757e-05,443 "loss": 1.5753582000732422,444 "step": 310445 },446 {447 "epoch": 0.340080971659919,448 "grad_norm": 3.59375,449 "learning_rate": 1.1555808282681187e-05,450 "loss": 1.3877858161926269,451 "step": 315452 },453 {454 "epoch": 0.34547908232118757,455 "grad_norm": 2.4375,456 "learning_rate": 1.1444963730971009e-05,457 "loss": 1.3945396423339844,458 "step": 320459 },460 {461 "epoch": 0.3508771929824561,462 "grad_norm": 3.125,463 "learning_rate": 1.1332914830316416e-05,464 "loss": 1.3506329536437989,465 "step": 325466 },467 {468 "epoch": 0.3562753036437247,469 "grad_norm": 3.296875,470 "learning_rate": 1.121969578786973e-05,471 "loss": 1.4695911407470703,472 "step": 330473 },474 {475 "epoch": 0.3616734143049933,476 "grad_norm": 4.9375,477 "learning_rate": 1.1105341168013172e-05,478 "loss": 1.350149440765381,479 "step": 335480 },481 {482 "epoch": 0.36707152496626183,483 "grad_norm": 2.859375,484 "learning_rate": 1.098988588180679e-05,485 "loss": 1.4387104034423828,486 "step": 340487 },488 {489 "epoch": 0.3724696356275304,490 "grad_norm": 3.421875,491 "learning_rate": 1.0873365176330553e-05,492 "loss": 1.3866162300109863,493 "step": 345494 },495 {496 "epoch": 0.37786774628879893,497 "grad_norm": 3.0625,498 "learning_rate": 1.0755814623923825e-05,499 "loss": 1.4232500076293946,500 "step": 350501 },502 {503 "epoch": 0.3832658569500675,504 "grad_norm": 2.890625,505 "learning_rate": 1.0637270111325591e-05,506 "loss": 1.483820629119873,507 "step": 355508 },509 {510 "epoch": 0.38866396761133604,511 "grad_norm": 2.40625,512 "learning_rate": 1.0517767828718664e-05,513 "loss": 1.429130458831787,514 "step": 360515 },516 {517 "epoch": 0.3940620782726046,518 "grad_norm": 3.1875,519 "learning_rate": 1.0397344258681267e-05,520 "loss": 1.435228157043457,521 "step": 365522 },523 {524 "epoch": 0.39946018893387314,525 "grad_norm": 3.328125,526 "learning_rate": 1.027603616504936e-05,527 "loss": 1.508815097808838,528 "step": 370529 },530 {531 "epoch": 0.4048582995951417,532 "grad_norm": 3.625,533 "learning_rate": 1.0153880581693075e-05,534 "loss": 1.486521339416504,535 "step": 375536 },537 {538 "epoch": 0.41025641025641024,539 "grad_norm": 4.09375,540 "learning_rate": 1.0030914801210753e-05,541 "loss": 1.3994467735290528,542 "step": 380543 },544 {545 "epoch": 0.4156545209176788,546 "grad_norm": 2.71875,547 "learning_rate": 9.907176363543956e-06,548 "loss": 1.417519474029541,549 "step": 385550 },551 {552 "epoch": 0.42105263157894735,553 "grad_norm": 3.203125,554 "learning_rate": 9.782703044516986e-06,555 "loss": 1.485072708129883,556 "step": 390557 },558 {559 "epoch": 0.42645074224021595,560 "grad_norm": 3.984375,561 "learning_rate": 9.657532844304396e-06,562 "loss": 1.4095831871032716,563 "step": 395564 },565 {566 "epoch": 0.4318488529014845,567 "grad_norm": 2.359375,568 "learning_rate": 9.531703975829994e-06,569 "loss": 1.5393587112426759,570 "step": 400571 },572 {573 "epoch": 0.43724696356275305,574 "grad_norm": 3.3125,575 "learning_rate": 9.405254853100927e-06,576 "loss": 1.4214500427246093,577 "step": 405578 },579 {580 "epoch": 0.4426450742240216,581 "grad_norm": 3.03125,582 "learning_rate": 9.278224079480331e-06,583 "loss": 1.3928979873657226,584 "step": 410585 },586 {587 "epoch": 0.44804318488529016,588 "grad_norm": 3.0625,589 "learning_rate": 9.150650435902215e-06,590 "loss": 1.4364460945129394,591 "step": 415592 },593 {594 "epoch": 0.4534412955465587,595 "grad_norm": 4.65625,596 "learning_rate": 9.022572869032115e-06,597 "loss": 1.3835034370422363,598 "step": 420599 },600 {601 "epoch": 0.45883940620782726,602 "grad_norm": 2.75,603 "learning_rate": 8.894030479377152e-06,604 "loss": 1.4032472610473632,605 "step": 425606 },607 {608 "epoch": 0.4642375168690958,609 "grad_norm": 3.84375,610 "learning_rate": 8.765062509349134e-06,611 "loss": 1.3748113632202148,612 "step": 430613 },614 {615 "epoch": 0.46963562753036436,616 "grad_norm": 2.78125,617 "learning_rate": 8.635708331284326e-06,618 "loss": 1.4337423324584961,619 "step": 435620 },621 {622 "epoch": 0.4750337381916329,623 "grad_norm": 3.015625,624 "learning_rate": 8.506007435423558e-06,625 "loss": 1.4276753425598145,626 "step": 440627 },628 {629 "epoch": 0.48043184885290147,630 "grad_norm": 2.375,631 "learning_rate": 8.375999417856343e-06,632 "loss": 1.3890512466430665,633 "step": 445634 },635 {636 "epoch": 0.48582995951417,637 "grad_norm": 2.875,638 "learning_rate": 8.24572396843268e-06,639 "loss": 1.4220175743103027,640 "step": 450641 },642 {643 "epoch": 0.49122807017543857,644 "grad_norm": 2.78125,645 "learning_rate": 8.11522085864622e-06,646 "loss": 1.4079001426696778,647 "step": 455648 },649 {650 "epoch": 0.4966261808367072,651 "grad_norm": 3.765625,652 "learning_rate": 7.984529929492535e-06,653 "loss": 1.369932270050049,654 "step": 460655 },656 {657 "epoch": 0.5020242914979757,658 "grad_norm": 2.578125,659 "learning_rate": 7.853691079306129e-06,660 "loss": 1.269005298614502,661 "step": 465662 },663 {664 "epoch": 0.5074224021592443,665 "grad_norm": 2.984375,666 "learning_rate": 7.722744251579988e-06,667 "loss": 1.3606270790100097,668 "step": 470669 },670 {671 "epoch": 0.5128205128205128,672 "grad_norm": 3.015625,673 "learning_rate": 7.591729422771317e-06,674 "loss": 1.4284911155700684,675 "step": 475676 },677 {678 "epoch": 0.5182186234817814,679 "grad_norm": 3.15625,680 "learning_rate": 7.460686590097202e-06,681 "loss": 1.449836254119873,682 "step": 480683 },684 {685 "epoch": 0.5236167341430499,686 "grad_norm": 4.21875,687 "learning_rate": 7.3296557593239795e-06,688 "loss": 1.3596923828125,689 "step": 485690 },691 {692 "epoch": 0.5290148448043185,693 "grad_norm": 2.796875,694 "learning_rate": 7.1986769325539485e-06,695 "loss": 1.3217558860778809,696 "step": 490697 },698 {699 "epoch": 0.5344129554655871,700 "grad_norm": 4.21875,701 "learning_rate": 7.067790096013219e-06,702 "loss": 1.4159544944763183,703 "step": 495704 },705 {706 "epoch": 0.5398110661268556,707 "grad_norm": 2.75,708 "learning_rate": 6.93703520784442e-06,709 "loss": 1.4698023796081543,710 "step": 500711 },712 {713 "epoch": 0.5452091767881242,714 "grad_norm": 2.59375,715 "learning_rate": 6.806452185907963e-06,716 "loss": 1.4592896461486817,717 "step": 505718 },719 {720 "epoch": 0.5506072874493927,721 "grad_norm": 2.578125,722 "learning_rate": 6.676080895595617e-06,723 "loss": 1.440026092529297,724 "step": 510725 },726 {727 "epoch": 0.5560053981106613,728 "grad_norm": 2.421875,729 "learning_rate": 6.545961137660077e-06,730 "loss": 1.4954944610595704,731 "step": 515732 },733 {734 "epoch": 0.5614035087719298,735 "grad_norm": 3.453125,736 "learning_rate": 6.416132636064301e-06,737 "loss": 1.4568126678466797,738 "step": 520739 },740 {741 "epoch": 0.5668016194331984,742 "grad_norm": 2.828125,743 "learning_rate": 6.2866350258542645e-06,744 "loss": 1.4184711456298829,745 "step": 525746 },747 {748 "epoch": 0.5721997300944669,749 "grad_norm": 2.890625,750 "learning_rate": 6.157507841058874e-06,751 "loss": 1.4063173294067384,752 "step": 530753 },754 {755 "epoch": 0.5775978407557355,756 "grad_norm": 3.15625,757 "learning_rate": 6.028790502620701e-06,758 "loss": 1.3004539489746094,759 "step": 535760 },761 {762 "epoch": 0.582995951417004,763 "grad_norm": 3.296875,764 "learning_rate": 5.900522306361271e-06,765 "loss": 1.3872806549072265,766 "step": 540767 },768 {769 "epoch": 0.5883940620782726,770 "grad_norm": 3.03125,771 "learning_rate": 5.772742410984511e-06,772 "loss": 1.3209671020507812,773 "step": 545774 },775 {776 "epoch": 0.5937921727395412,777 "grad_norm": 2.640625,778 "learning_rate": 5.645489826122098e-06,779 "loss": 1.5138866424560546,780 "step": 550781 },782 {783 "epoch": 0.5991902834008097,784 "grad_norm": 3.75,785 "learning_rate": 5.518803400424286e-06,786 "loss": 1.383652400970459,787 "step": 555788 },789 {790 "epoch": 0.6045883940620783,791 "grad_norm": 3.625,792 "learning_rate": 5.392721809699897e-06,793 "loss": 1.4016658782958984,794 "step": 560795 },796 {797 "epoch": 0.6099865047233468,798 "grad_norm": 2.796875,799 "learning_rate": 5.267283545109068e-06,800 "loss": 1.3988656997680664,801 "step": 565802 },803 {804 "epoch": 0.6153846153846154,805 "grad_norm": 2.65625,806 "learning_rate": 5.142526901412366e-06,807 "loss": 1.4131988525390624,808 "step": 570809 },810 {811 "epoch": 0.6207827260458839,812 "grad_norm": 3.5625,813 "learning_rate": 5.018489965279887e-06,814 "loss": 1.3934114456176758,815 "step": 575816 },817 {818 "epoch": 0.6261808367071525,819 "grad_norm": 3.09375,820 "learning_rate": 4.895210603663849e-06,821 "loss": 1.3578675270080567,822 "step": 580823 },824 {825 "epoch": 0.631578947368421,826 "grad_norm": 3.015625,827 "learning_rate": 4.772726452238277e-06,828 "loss": 1.4002360343933105,829 "step": 585830 },831 {832 "epoch": 0.6369770580296896,833 "grad_norm": 4.21875,834 "learning_rate": 4.651074903909303e-06,835 "loss": 1.2647083282470704,836 "step": 590837 },838 {839 "epoch": 0.6423751686909581,840 "grad_norm": 2.984375,841 "learning_rate": 4.5302930973995636e-06,842 "loss": 1.402871036529541,843 "step": 595844 },845 {846 "epoch": 0.6477732793522267,847 "grad_norm": 2.25,848 "learning_rate": 4.4104179059102066e-06,849 "loss": 1.3439860343933105,850 "step": 600851 },852 {853 "epoch": 0.6531713900134952,854 "grad_norm": 3.796875,855 "learning_rate": 4.291485925863952e-06,856 "loss": 1.3316984176635742,857 "step": 605858 },859 {860 "epoch": 0.6585695006747638,861 "grad_norm": 2.875,862 "learning_rate": 4.173533465732669e-06,863 "loss": 1.4628570556640625,864 "step": 610865 },866 {867 "epoch": 0.6639676113360324,868 "grad_norm": 3.390625,869 "learning_rate": 4.056596534952823e-06,870 "loss": 1.3681221961975099,871 "step": 615872 },873 {874 "epoch": 0.6693657219973009,875 "grad_norm": 2.5625,876 "learning_rate": 3.940710832932262e-06,877 "loss": 1.4698673248291017,878 "step": 620879 },880 {881 "epoch": 0.6747638326585695,882 "grad_norm": 3.09375,883 "learning_rate": 3.8259117381516e-06,884 "loss": 1.3489319801330566,885 "step": 625886 },887 {888 "epoch": 0.680161943319838,889 "grad_norm": 2.734375,890 "learning_rate": 3.712234297363618e-06,891 "loss": 1.3783289909362793,892 "step": 630893 },894 {895 "epoch": 0.6855600539811066,896 "grad_norm": 2.78125,897 "learning_rate": 3.5997132148939237e-06,898 "loss": 1.5137944221496582,899 "step": 635900 },901 {902 "epoch": 0.6909581646423751,903 "grad_norm": 4.46875,904 "learning_rate": 3.488382842046137e-06,905 "loss": 1.2776326179504394,906 "step": 640907 },908 {909 "epoch": 0.6963562753036437,910 "grad_norm": 3.953125,911 "learning_rate": 3.3782771666148696e-06,912 "loss": 1.3425775527954102,913 "step": 645914 },915 {916 "epoch": 0.7017543859649122,917 "grad_norm": 3.28125,918 "learning_rate": 3.26942980250965e-06,919 "loss": 1.3643362998962403,920 "step": 650921 },922 {923 "epoch": 0.7071524966261808,924 "grad_norm": 2.78125,925 "learning_rate": 3.1618739794930366e-06,926 "loss": 1.4447607040405273,927 "step": 655928 },929 {930 "epoch": 0.7125506072874493,931 "grad_norm": 2.828125,932 "learning_rate": 3.055642533035964e-06,933 "loss": 1.406121826171875,934 "step": 660935 },936 {937 "epoch": 0.717948717948718,938 "grad_norm": 3.765625,939 "learning_rate": 2.9507678942934702e-06,940 "loss": 1.2962283134460448,941 "step": 665942 },943 {944 "epoch": 0.7233468286099866,945 "grad_norm": 3.859375,946 "learning_rate": 2.8472820802038847e-06,947 "loss": 1.3783414840698243,948 "step": 670949 },950 {951 "epoch": 0.728744939271255,952 "grad_norm": 4.1875,953 "learning_rate": 2.745216683714428e-06,954 "loss": 1.3261849403381347,955 "step": 675956 },957 {958 "epoch": 0.7341430499325237,959 "grad_norm": 2.96875,960 "learning_rate": 2.6446028641362934e-06,961 "loss": 1.5042543411254883,962 "step": 680963 },964 {965 "epoch": 0.7395411605937922,966 "grad_norm": 3.875,967 "learning_rate": 2.5454713376320764e-06,968 "loss": 1.3572941780090333,969 "step": 685970 },971 {972 "epoch": 0.7449392712550608,973 "grad_norm": 3.046875,974 "learning_rate": 2.4478523678385194e-06,975 "loss": 1.3866613388061524,976 "step": 690977 },978 {979 "epoch": 0.7503373819163293,980 "grad_norm": 2.90625,981 "learning_rate": 2.3517757566273827e-06,982 "loss": 1.4149866104125977,983 "step": 695984 },985 {986 "epoch": 0.7557354925775979,987 "grad_norm": 3.078125,988 "learning_rate": 2.257270835007314e-06,989 "loss": 1.3730114936828612,990 "step": 700991 },992 {993 "epoch": 0.7611336032388664,994 "grad_norm": 4.40625,995 "learning_rate": 2.1643664541694376e-06,996 "loss": 1.2785679817199707,997 "step": 705998 },999 {1000 "epoch": 0.766531713900135,1001 "grad_norm": 2.671875,1002 "learning_rate": 2.073090976679452e-06,1003 "loss": 1.229942226409912,1004 "step": 7101005 },1006 {1007 "epoch": 0.7719298245614035,1008 "grad_norm": 2.375,1009 "learning_rate": 1.983472267818883e-06,1010 "loss": 1.3718143463134767,1011 "step": 7151012 },1013 {1014 "epoch": 0.7773279352226721,1015 "grad_norm": 2.359375,1016 "learning_rate": 1.8955376870781552e-06,1017 "loss": 1.402730655670166,1018 "step": 7201019 },1020 {1021 "epoch": 0.7827260458839406,1022 "grad_norm": 3.390625,1023 "learning_rate": 1.8093140798040686e-06,1024 "loss": 1.3430524826049806,1025 "step": 7251026 },1027 {1028 "epoch": 0.7881241565452092,1029 "grad_norm": 2.3125,1030 "learning_rate": 1.7248277690042604e-06,1031 "loss": 1.4013592720031738,1032 "step": 7301033 },1034 {1035 "epoch": 0.7935222672064778,1036 "grad_norm": 2.40625,1037 "learning_rate": 1.642104547311099e-06,1038 "loss": 1.407256507873535,1039 "step": 7351040 },1041 {1042 "epoch": 0.7989203778677463,1043 "grad_norm": 2.171875,1044 "learning_rate": 1.5611696691075208e-06,1045 "loss": 1.4018988609313965,1046 "step": 7401047 },1048 {1049 "epoch": 0.8043184885290149,1050 "grad_norm": 2.703125,1051 "learning_rate": 1.4820478428171713e-06,1052 "loss": 1.3888368606567383,1053 "step": 7451054 },1055 {1056 "epoch": 0.8097165991902834,1057 "grad_norm": 2.125,1058 "learning_rate": 1.4047632233612263e-06,1059 "loss": 1.444268035888672,1060 "step": 7501061 },1062 {1063 "epoch": 0.815114709851552,1064 "grad_norm": 3.1875,1065 "learning_rate": 1.3293394047841987e-06,1066 "loss": 1.4442663192749023,1067 "step": 7551068 },1069 {1070 "epoch": 0.8205128205128205,1071 "grad_norm": 2.328125,1072 "learning_rate": 1.2557994130509583e-06,1073 "loss": 1.4278512001037598,1074 "step": 7601075 },1076 {1077 "epoch": 0.8259109311740891,1078 "grad_norm": 3.421875,1079 "learning_rate": 1.184165699017195e-06,1080 "loss": 1.4104692459106445,1081 "step": 7651082 },1083 {1084 "epoch": 0.8313090418353576,1085 "grad_norm": 3.96875,1086 "learning_rate": 1.1144601315754397e-06,1087 "loss": 1.3686716079711914,1088 "step": 7701089 },1090 {1091 "epoch": 0.8367071524966262,1092 "grad_norm": 2.53125,1093 "learning_rate": 1.0467039909787668e-06,1094 "loss": 1.4443007469177247,1095 "step": 7751096 },1097 {1098 "epoch": 0.8421052631578947,1099 "grad_norm": 4.125,1100 "learning_rate": 9.809179623441958e-07,1101 "loss": 1.411851978302002,1102 "step": 7801103 },1104 {1105 "epoch": 0.8475033738191633,1106 "grad_norm": 2.328125,1107 "learning_rate": 9.171221293377735e-07,1108 "loss": 1.491429901123047,1109 "step": 7851110 },1111 {1112 "epoch": 0.8529014844804319,1113 "grad_norm": 2.53125,1114 "learning_rate": 8.553359680432815e-07,1115 "loss": 1.3643847465515138,1116 "step": 7901117 },1118 {1119 "epoch": 0.8582995951417004,1120 "grad_norm": 3.5,1121 "learning_rate": 7.95578341016428e-07,1122 "loss": 1.3733964920043946,1123 "step": 7951124 },1125 {1126 "epoch": 0.863697705802969,1127 "grad_norm": 4.375,1128 "learning_rate": 7.378674915263359e-07,1129 "loss": 1.270117950439453,1130 "step": 8001131 },1132 {1133 "epoch": 0.8690958164642375,1134 "grad_norm": 3.03125,1135 "learning_rate": 6.822210379861002e-07,1136 "loss": 1.4278650283813477,1137 "step": 8051138 },1139 {1140 "epoch": 0.8744939271255061,1141 "grad_norm": 2.671875,1142 "learning_rate": 6.286559685741076e-07,1143 "loss": 1.388138484954834,1144 "step": 8101145 },1146 {1147 "epoch": 0.8798920377867746,1148 "grad_norm": 3.5,1149 "learning_rate": 5.77188636047746e-07,1150 "loss": 1.5160203933715821,1151 "step": 8151152 },1153 {1154 "epoch": 0.8852901484480432,1155 "grad_norm": 3.21875,1156 "learning_rate": 5.278347527511174e-07,1157 "loss": 1.3769659042358398,1158 "step": 8201159 },1160 {1161 "epoch": 0.8906882591093117,1162 "grad_norm": 2.859375,1163 "learning_rate": 4.806093858182428e-07,1164 "loss": 1.4992185592651368,1165 "step": 8251166 },1167 {1168 "epoch": 0.8960863697705803,1169 "grad_norm": 3.21875,1170 "learning_rate": 4.3552695257325775e-07,1171 "loss": 1.3635730743408203,1172 "step": 8301173 },1174 {1175 "epoch": 0.9014844804318488,1176 "grad_norm": 2.09375,1177 "learning_rate": 3.9260121612898244e-07,1178 "loss": 1.4149579048156737,1179 "step": 8351180 },1181 {1182 "epoch": 0.9068825910931174,1183 "grad_norm": 4.09375,1184 "learning_rate": 3.5184528118520296e-07,1185 "loss": 1.4671350479125977,1186 "step": 8401187 },1188 {1189 "epoch": 0.9122807017543859,1190 "grad_norm": 2.640625,1191 "learning_rate": 3.132715900279809e-07,1192 "loss": 1.413409423828125,1193 "step": 8451194 },1195 {1196 "epoch": 0.9176788124156545,1197 "grad_norm": 3.015625,1198 "learning_rate": 2.768919187311614e-07,1199 "loss": 1.399053192138672,1200 "step": 850