windgrin/q3
02
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 2.9834983498349836,6 "eval_steps": 500,7 "global_step": 453,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.006600660066006601,14 "grad_norm": 3.269663095474243,15 "learning_rate": 0.0,16 "loss": 2.0173,17 "step": 118 },19 {20 "epoch": 0.013201320132013201,21 "grad_norm": 3.4657950401306152,22 "learning_rate": 4.347826086956522e-06,23 "loss": 1.8773,24 "step": 225 },26 {27 "epoch": 0.019801980198019802,28 "grad_norm": 3.480679512023926,29 "learning_rate": 8.695652173913044e-06,30 "loss": 2.2683,31 "step": 332 },33 {34 "epoch": 0.026402640264026403,35 "grad_norm": 3.3081204891204834,36 "learning_rate": 1.3043478260869566e-05,37 "loss": 1.5357,38 "step": 439 },40 {41 "epoch": 0.033003300330033,42 "grad_norm": 2.9084601402282715,43 "learning_rate": 1.739130434782609e-05,44 "loss": 2.3349,45 "step": 546 },47 {48 "epoch": 0.039603960396039604,49 "grad_norm": 4.738473892211914,50 "learning_rate": 2.173913043478261e-05,51 "loss": 2.0195,52 "step": 653 },54 {55 "epoch": 0.0462046204620462,56 "grad_norm": 3.0607786178588867,57 "learning_rate": 2.608695652173913e-05,58 "loss": 1.7303,59 "step": 760 },61 {62 "epoch": 0.052805280528052806,63 "grad_norm": 3.43908429145813,64 "learning_rate": 3.0434782608695656e-05,65 "loss": 1.5714,66 "step": 867 },68 {69 "epoch": 0.0594059405940594,70 "grad_norm": 3.3616912364959717,71 "learning_rate": 3.478260869565218e-05,72 "loss": 2.1655,73 "step": 974 },75 {76 "epoch": 0.066006600660066,77 "grad_norm": 2.8642077445983887,78 "learning_rate": 3.91304347826087e-05,79 "loss": 2.2225,80 "step": 1081 },82 {83 "epoch": 0.07260726072607261,84 "grad_norm": 1.762931227684021,85 "learning_rate": 4.347826086956522e-05,86 "loss": 4.3722,87 "step": 1188 },89 {90 "epoch": 0.07920792079207921,91 "grad_norm": 2.3837015628814697,92 "learning_rate": 4.782608695652174e-05,93 "loss": 1.134,94 "step": 1295 },96 {97 "epoch": 0.0858085808580858,98 "grad_norm": 3.5527191162109375,99 "learning_rate": 5.217391304347826e-05,100 "loss": 1.5813,101 "step": 13102 },103 {104 "epoch": 0.0924092409240924,105 "grad_norm": 4.409274578094482,106 "learning_rate": 5.652173913043478e-05,107 "loss": 1.8402,108 "step": 14109 },110 {111 "epoch": 0.09900990099009901,112 "grad_norm": 3.351221799850464,113 "learning_rate": 6.086956521739131e-05,114 "loss": 1.6236,115 "step": 15116 },117 {118 "epoch": 0.10561056105610561,119 "grad_norm": 3.299344301223755,120 "learning_rate": 6.521739130434783e-05,121 "loss": 1.4048,122 "step": 16123 },124 {125 "epoch": 0.11221122112211221,126 "grad_norm": 2.9774715900421143,127 "learning_rate": 6.956521739130436e-05,128 "loss": 1.5322,129 "step": 17130 },131 {132 "epoch": 0.1188118811881188,133 "grad_norm": 3.526827096939087,134 "learning_rate": 7.391304347826086e-05,135 "loss": 1.2128,136 "step": 18137 },138 {139 "epoch": 0.1254125412541254,140 "grad_norm": 1.8688900470733643,141 "learning_rate": 7.82608695652174e-05,142 "loss": 1.1241,143 "step": 19144 },145 {146 "epoch": 0.132013201320132,147 "grad_norm": 2.7325925827026367,148 "learning_rate": 8.260869565217392e-05,149 "loss": 1.4779,150 "step": 20151 },152 {153 "epoch": 0.13861386138613863,154 "grad_norm": 2.606200695037842,155 "learning_rate": 8.695652173913044e-05,156 "loss": 1.2094,157 "step": 21158 },159 {160 "epoch": 0.14521452145214522,161 "grad_norm": 3.906005859375,162 "learning_rate": 9.130434782608696e-05,163 "loss": 1.9388,164 "step": 22165 },166 {167 "epoch": 0.15181518151815182,168 "grad_norm": 3.1924426555633545,169 "learning_rate": 9.565217391304348e-05,170 "loss": 1.6678,171 "step": 23172 },173 {174 "epoch": 0.15841584158415842,175 "grad_norm": 3.2093305587768555,176 "learning_rate": 0.0001,177 "loss": 1.8155,178 "step": 24179 },180 {181 "epoch": 0.16501650165016502,182 "grad_norm": 2.6762614250183105,183 "learning_rate": 0.00010434782608695653,184 "loss": 1.8512,185 "step": 25186 },187 {188 "epoch": 0.1716171617161716,189 "grad_norm": 2.4602980613708496,190 "learning_rate": 0.00010869565217391305,191 "loss": 1.4496,192 "step": 26193 },194 {195 "epoch": 0.1782178217821782,196 "grad_norm": 2.621077299118042,197 "learning_rate": 0.00011304347826086956,198 "loss": 1.7004,199 "step": 27200 },201 {202 "epoch": 0.1848184818481848,203 "grad_norm": 3.040069580078125,204 "learning_rate": 0.0001173913043478261,205 "loss": 1.3742,206 "step": 28207 },208 {209 "epoch": 0.19141914191419143,210 "grad_norm": 2.6296839714050293,211 "learning_rate": 0.00012173913043478263,212 "loss": 1.9487,213 "step": 29214 },215 {216 "epoch": 0.19801980198019803,217 "grad_norm": 2.5183770656585693,218 "learning_rate": 0.00012608695652173915,219 "loss": 1.4286,220 "step": 30221 },222 {223 "epoch": 0.20462046204620463,224 "grad_norm": 2.7780044078826904,225 "learning_rate": 0.00013043478260869567,226 "loss": 1.6399,227 "step": 31228 },229 {230 "epoch": 0.21122112211221122,231 "grad_norm": 4.337399959564209,232 "learning_rate": 0.0001347826086956522,233 "loss": 1.8436,234 "step": 32235 },236 {237 "epoch": 0.21782178217821782,238 "grad_norm": 2.1603963375091553,239 "learning_rate": 0.0001391304347826087,240 "loss": 1.3298,241 "step": 33242 },243 {244 "epoch": 0.22442244224422442,245 "grad_norm": 2.3995110988616943,246 "learning_rate": 0.0001434782608695652,247 "loss": 1.4131,248 "step": 34249 },250 {251 "epoch": 0.23102310231023102,252 "grad_norm": 2.0356786251068115,253 "learning_rate": 0.00014782608695652173,254 "loss": 1.2884,255 "step": 35256 },257 {258 "epoch": 0.2376237623762376,259 "grad_norm": 2.234151840209961,260 "learning_rate": 0.00015217391304347827,261 "loss": 1.3331,262 "step": 36263 },264 {265 "epoch": 0.24422442244224424,266 "grad_norm": 2.045015573501587,267 "learning_rate": 0.0001565217391304348,268 "loss": 1.1223,269 "step": 37270 },271 {272 "epoch": 0.2508250825082508,273 "grad_norm": 2.8687806129455566,274 "learning_rate": 0.00016086956521739132,275 "loss": 1.3582,276 "step": 38277 },278 {279 "epoch": 0.25742574257425743,280 "grad_norm": 2.717107057571411,281 "learning_rate": 0.00016521739130434784,282 "loss": 1.8503,283 "step": 39284 },285 {286 "epoch": 0.264026402640264,287 "grad_norm": 2.2588422298431396,288 "learning_rate": 0.00016956521739130436,289 "loss": 1.4314,290 "step": 40291 },292 {293 "epoch": 0.2706270627062706,294 "grad_norm": 2.3839213848114014,295 "learning_rate": 0.00017391304347826088,296 "loss": 1.6702,297 "step": 41298 },299 {300 "epoch": 0.27722772277227725,301 "grad_norm": 2.2554593086242676,302 "learning_rate": 0.0001782608695652174,303 "loss": 1.6114,304 "step": 42305 },306 {307 "epoch": 0.2838283828382838,308 "grad_norm": 3.0558981895446777,309 "learning_rate": 0.00018260869565217392,310 "loss": 1.7576,311 "step": 43312 },313 {314 "epoch": 0.29042904290429045,315 "grad_norm": 1.9171123504638672,316 "learning_rate": 0.00018695652173913045,317 "loss": 1.0488,318 "step": 44319 },320 {321 "epoch": 0.297029702970297,322 "grad_norm": 2.573887825012207,323 "learning_rate": 0.00019130434782608697,324 "loss": 1.5062,325 "step": 45326 },327 {328 "epoch": 0.30363036303630364,329 "grad_norm": 2.058779716491699,330 "learning_rate": 0.0001956521739130435,331 "loss": 1.1712,332 "step": 46333 },334 {335 "epoch": 0.3102310231023102,336 "grad_norm": 4.420455455780029,337 "learning_rate": 0.0002,338 "loss": 2.2398,339 "step": 47340 },341 {342 "epoch": 0.31683168316831684,343 "grad_norm": 2.39105224609375,344 "learning_rate": 0.00019999702094326033,345 "loss": 1.3762,346 "step": 48347 },348 {349 "epoch": 0.3234323432343234,350 "grad_norm": 1.9990428686141968,351 "learning_rate": 0.00019998808395053688,352 "loss": 1.8914,353 "step": 49354 },355 {356 "epoch": 0.33003300330033003,357 "grad_norm": 1.8856112957000732,358 "learning_rate": 0.0001999731895543058,359 "loss": 1.239,360 "step": 50361 },362 {363 "epoch": 0.33663366336633666,364 "grad_norm": 1.7601982355117798,365 "learning_rate": 0.00019995233864199215,366 "loss": 1.31,367 "step": 51368 },369 {370 "epoch": 0.3432343234323432,371 "grad_norm": 2.37358021736145,372 "learning_rate": 0.00019992553245591694,373 "loss": 1.2271,374 "step": 52375 },376 {377 "epoch": 0.34983498349834985,378 "grad_norm": 1.8610317707061768,379 "learning_rate": 0.00019989277259322314,380 "loss": 1.222,381 "step": 53382 },383 {384 "epoch": 0.3564356435643564,385 "grad_norm": 2.4520764350891113,386 "learning_rate": 0.0001998540610057806,387 "loss": 1.8557,388 "step": 54389 },390 {391 "epoch": 0.36303630363036304,392 "grad_norm": 1.9622912406921387,393 "learning_rate": 0.00019980940000006954,394 "loss": 1.3294,395 "step": 55396 },397 {398 "epoch": 0.3696369636963696,399 "grad_norm": 2.875389337539673,400 "learning_rate": 0.0001997587922370434,401 "loss": 1.0667,402 "step": 56403 },404 {405 "epoch": 0.37623762376237624,406 "grad_norm": 2.358046531677246,407 "learning_rate": 0.00019970224073197017,408 "loss": 2.0223,409 "step": 57410 },411 {412 "epoch": 0.38283828382838286,413 "grad_norm": 2.127352237701416,414 "learning_rate": 0.00019963974885425266,415 "loss": 1.5623,416 "step": 58417 },418 {419 "epoch": 0.38943894389438943,420 "grad_norm": 2.165083885192871,421 "learning_rate": 0.00019957132032722785,422 "loss": 1.6115,423 "step": 59424 },425 {426 "epoch": 0.39603960396039606,427 "grad_norm": 2.4137611389160156,428 "learning_rate": 0.00019949695922794506,429 "loss": 1.5646,430 "step": 60431 },432 {433 "epoch": 0.40264026402640263,434 "grad_norm": 2.246424913406372,435 "learning_rate": 0.00019941666998692296,436 "loss": 1.7207,437 "step": 61438 },439 {440 "epoch": 0.40924092409240925,441 "grad_norm": 2.1227519512176514,442 "learning_rate": 0.00019933045738788563,443 "loss": 1.6483,444 "step": 62445 },446 {447 "epoch": 0.4158415841584158,448 "grad_norm": 2.097881317138672,449 "learning_rate": 0.0001992383265674776,450 "loss": 1.2604,451 "step": 63452 },453 {454 "epoch": 0.42244224422442245,455 "grad_norm": 2.4792444705963135,456 "learning_rate": 0.0001991402830149576,457 "loss": 1.3448,458 "step": 64459 },460 {461 "epoch": 0.429042904290429,462 "grad_norm": 2.302548408508301,463 "learning_rate": 0.00019903633257187185,464 "loss": 1.5288,465 "step": 65466 },467 {468 "epoch": 0.43564356435643564,469 "grad_norm": 1.9828969240188599,470 "learning_rate": 0.00019892648143170566,471 "loss": 1.1426,472 "step": 66473 },474 {475 "epoch": 0.44224422442244227,476 "grad_norm": 2.2925431728363037,477 "learning_rate": 0.0001988107361395146,478 "loss": 1.3227,479 "step": 67480 },481 {482 "epoch": 0.44884488448844884,483 "grad_norm": 3.0168843269348145,484 "learning_rate": 0.0001986891035915346,485 "loss": 2.0713,486 "step": 68487 },488 {489 "epoch": 0.45544554455445546,490 "grad_norm": 2.2189230918884277,491 "learning_rate": 0.00019856159103477086,492 "loss": 1.2472,493 "step": 69494 },495 {496 "epoch": 0.46204620462046203,497 "grad_norm": 1.9616774320602417,498 "learning_rate": 0.0001984282060665662,499 "loss": 1.4617,500 "step": 70501 },502 {503 "epoch": 0.46864686468646866,504 "grad_norm": 3.0113203525543213,505 "learning_rate": 0.00019828895663414836,506 "loss": 1.5551,507 "step": 71508 },509 {510 "epoch": 0.4752475247524752,511 "grad_norm": 2.4521045684814453,512 "learning_rate": 0.00019814385103415663,513 "loss": 1.405,514 "step": 72515 },516 {517 "epoch": 0.48184818481848185,518 "grad_norm": 2.3250396251678467,519 "learning_rate": 0.00019799289791214725,520 "loss": 1.8011,521 "step": 73522 },523 {524 "epoch": 0.4884488448844885,525 "grad_norm": 2.3330326080322266,526 "learning_rate": 0.00019783610626207854,527 "loss": 1.6456,528 "step": 74529 },530 {531 "epoch": 0.49504950495049505,532 "grad_norm": 2.1660053730010986,533 "learning_rate": 0.00019767348542577495,534 "loss": 1.5428,535 "step": 75536 },537 {538 "epoch": 0.5016501650165016,539 "grad_norm": 1.8240512609481812,540 "learning_rate": 0.00019750504509237046,541 "loss": 1.4666,542 "step": 76543 },544 {545 "epoch": 0.5082508250825083,546 "grad_norm": 2.2784554958343506,547 "learning_rate": 0.00019733079529773125,548 "loss": 1.5065,549 "step": 77550 },551 {552 "epoch": 0.5148514851485149,553 "grad_norm": 1.769106149673462,554 "learning_rate": 0.00019715074642385786,555 "loss": 1.0998,556 "step": 78557 },558 {559 "epoch": 0.5214521452145214,560 "grad_norm": 1.737535834312439,561 "learning_rate": 0.00019696490919826647,562 "loss": 1.3843,563 "step": 79564 },565 {566 "epoch": 0.528052805280528,567 "grad_norm": 1.737778663635254,568 "learning_rate": 0.0001967732946933499,569 "loss": 1.5803,570 "step": 80571 },572 {573 "epoch": 0.5346534653465347,574 "grad_norm": 1.9771817922592163,575 "learning_rate": 0.00019657591432571778,576 "loss": 1.7888,577 "step": 81578 },579 {580 "epoch": 0.5412541254125413,581 "grad_norm": 1.922725796699524,582 "learning_rate": 0.0001963727798555164,583 "loss": 1.096,584 "step": 82585 },586 {587 "epoch": 0.5478547854785478,588 "grad_norm": 2.42258620262146,589 "learning_rate": 0.00019616390338572803,590 "loss": 1.4231,591 "step": 83592 },593 {594 "epoch": 0.5544554455445545,595 "grad_norm": 2.0959455966949463,596 "learning_rate": 0.00019594929736144976,597 "loss": 1.3786,598 "step": 84599 },600 {601 "epoch": 0.5610561056105611,602 "grad_norm": 2.2483432292938232,603 "learning_rate": 0.00019572897456915202,604 "loss": 1.5456,605 "step": 85606 },607 {608 "epoch": 0.5676567656765676,609 "grad_norm": 2.2256951332092285,610 "learning_rate": 0.00019550294813591682,611 "loss": 1.3946,612 "step": 86613 },614 {615 "epoch": 0.5742574257425742,616 "grad_norm": 1.9461501836776733,617 "learning_rate": 0.0001952712315286556,618 "loss": 1.3132,619 "step": 87620 },621 {622 "epoch": 0.5808580858085809,623 "grad_norm": 1.8409348726272583,624 "learning_rate": 0.00019503383855330668,625 "loss": 1.1527,626 "step": 88627 },628 {629 "epoch": 0.5874587458745875,630 "grad_norm": 2.124138832092285,631 "learning_rate": 0.00019479078335401297,632 "loss": 1.0499,633 "step": 89634 },635 {636 "epoch": 0.594059405940594,637 "grad_norm": 2.380868673324585,638 "learning_rate": 0.00019454208041227907,639 "loss": 1.7297,640 "step": 90641 },642 {643 "epoch": 0.6006600660066007,644 "grad_norm": 2.093554973602295,645 "learning_rate": 0.00019428774454610843,646 "loss": 1.3311,647 "step": 91648 },649 {650 "epoch": 0.6072607260726073,651 "grad_norm": 2.391789436340332,652 "learning_rate": 0.0001940277909091206,653 "loss": 1.4725,654 "step": 92655 },656 {657 "epoch": 0.6138613861386139,658 "grad_norm": 2.362126350402832,659 "learning_rate": 0.0001937622349896483,660 "loss": 1.4505,661 "step": 93662 },663 {664 "epoch": 0.6204620462046204,665 "grad_norm": 2.4472386837005615,666 "learning_rate": 0.00019349109260981452,667 "loss": 1.4643,668 "step": 94669 },670 {671 "epoch": 0.6270627062706271,672 "grad_norm": 2.7160544395446777,673 "learning_rate": 0.00019321437992458995,674 "loss": 1.6129,675 "step": 95676 },677 {678 "epoch": 0.6336633663366337,679 "grad_norm": 1.9917261600494385,680 "learning_rate": 0.0001929321134208304,681 "loss": 1.4359,682 "step": 96683 },684 {685 "epoch": 0.6402640264026402,686 "grad_norm": 1.98976731300354,687 "learning_rate": 0.00019264430991629446,688 "loss": 1.4551,689 "step": 97690 },691 {692 "epoch": 0.6468646864686468,693 "grad_norm": 2.612914562225342,694 "learning_rate": 0.00019235098655864157,695 "loss": 1.4066,696 "step": 98697 },698 {699 "epoch": 0.6534653465346535,700 "grad_norm": 1.5987509489059448,701 "learning_rate": 0.00019205216082441017,702 "loss": 1.4239,703 "step": 99704 },705 {706 "epoch": 0.6600660066006601,707 "grad_norm": 2.1327996253967285,708 "learning_rate": 0.00019174785051797666,709 "loss": 1.1858,710 "step": 100711 },712 {713 "epoch": 0.6666666666666666,714 "grad_norm": 3.5038344860076904,715 "learning_rate": 0.0001914380737704944,716 "loss": 1.5437,717 "step": 101718 },719 {720 "epoch": 0.6732673267326733,721 "grad_norm": 2.0759711265563965,722 "learning_rate": 0.0001911228490388136,723 "loss": 1.9746,724 "step": 102725 },726 {727 "epoch": 0.6798679867986799,728 "grad_norm": 1.7349865436553955,729 "learning_rate": 0.0001908021951043814,730 "loss": 1.3087,731 "step": 103732 },733 {734 "epoch": 0.6864686468646864,735 "grad_norm": 1.8272370100021362,736 "learning_rate": 0.00019047613107212312,737 "loss": 1.5015,738 "step": 104739 },740 {741 "epoch": 0.693069306930693,742 "grad_norm": 2.6153130531311035,743 "learning_rate": 0.00019014467636930385,744 "loss": 1.6917,745 "step": 105746 },747 {748 "epoch": 0.6996699669966997,749 "grad_norm": 2.020885705947876,750 "learning_rate": 0.00018980785074437093,751 "loss": 1.3401,752 "step": 106753 },754 {755 "epoch": 0.7062706270627063,756 "grad_norm": 2.322054147720337,757 "learning_rate": 0.00018946567426577723,758 "loss": 2.2148,759 "step": 107760 },761 {762 "epoch": 0.7128712871287128,763 "grad_norm": 2.2210605144500732,764 "learning_rate": 0.00018911816732078575,765 "loss": 1.6508,766 "step": 108767 },768 {769 "epoch": 0.7194719471947195,770 "grad_norm": 2.340949058532715,771 "learning_rate": 0.00018876535061425453,772 "loss": 1.8924,773 "step": 109774 },775 {776 "epoch": 0.7260726072607261,777 "grad_norm": 2.295544385910034,778 "learning_rate": 0.0001884072451674034,779 "loss": 1.621,780 "step": 110781 },782 {783 "epoch": 0.7326732673267327,784 "grad_norm": 1.9485543966293335,785 "learning_rate": 0.00018804387231656118,786 "loss": 1.3108,787 "step": 111788 },789 {790 "epoch": 0.7392739273927392,791 "grad_norm": 1.9073518514633179,792 "learning_rate": 0.00018767525371189475,793 "loss": 1.1203,794 "step": 112795 },796 {797 "epoch": 0.7458745874587459,798 "grad_norm": 1.9978621006011963,799 "learning_rate": 0.00018730141131611882,800 "loss": 1.5631,801 "step": 113802 },803 {804 "epoch": 0.7524752475247525,805 "grad_norm": 2.061638355255127,806 "learning_rate": 0.0001869223674031876,807 "loss": 1.4612,808 "step": 114809 },810 {811 "epoch": 0.759075907590759,812 "grad_norm": 1.9964576959609985,813 "learning_rate": 0.00018653814455696756,814 "loss": 1.4753,815 "step": 115816 },817 {818 "epoch": 0.7656765676567657,819 "grad_norm": 2.142472505569458,820 "learning_rate": 0.0001861487656698919,821 "loss": 1.4195,822 "step": 116823 },824 {825 "epoch": 0.7722772277227723,826 "grad_norm": 2.622685194015503,827 "learning_rate": 0.00018575425394159653,828 "loss": 1.8376,829 "step": 117830 },831 {832 "epoch": 0.7788778877887789,833 "grad_norm": 1.8622280359268188,834 "learning_rate": 0.00018535463287753796,835 "loss": 1.3869,836 "step": 118837 },838 {839 "epoch": 0.7854785478547854,840 "grad_norm": 1.8239712715148926,841 "learning_rate": 0.00018494992628759267,842 "loss": 1.3901,843 "step": 119844 },845 {846 "epoch": 0.7920792079207921,847 "grad_norm": 1.7509407997131348,848 "learning_rate": 0.0001845401582846385,849 "loss": 1.1086,850 "step": 120851 },852 {853 "epoch": 0.7986798679867987,854 "grad_norm": 2.12126088142395,855 "learning_rate": 0.00018412535328311814,856 "loss": 1.4786,857 "step": 121858 },859 {860 "epoch": 0.8052805280528053,861 "grad_norm": 1.9442270994186401,862 "learning_rate": 0.00018370553599758422,863 "loss": 1.2655,864 "step": 122865 },866 {867 "epoch": 0.8118811881188119,868 "grad_norm": 1.8618887662887573,869 "learning_rate": 0.00018328073144122708,870 "loss": 1.1703,871 "step": 123872 },873 {874 "epoch": 0.8184818481848185,875 "grad_norm": 1.9630706310272217,876 "learning_rate": 0.00018285096492438424,877 "loss": 1.2309,878 "step": 124879 },880 {881 "epoch": 0.8250825082508251,882 "grad_norm": 2.227321147918701,883 "learning_rate": 0.00018241626205303243,884 "loss": 1.6171,885 "step": 125886 },887 {888 "epoch": 0.8316831683168316,889 "grad_norm": 1.6737253665924072,890 "learning_rate": 0.00018197664872726205,891 "loss": 1.0966,892 "step": 126893 },894 {895 "epoch": 0.8382838283828383,896 "grad_norm": 1.8567694425582886,897 "learning_rate": 0.00018153215113973396,898 "loss": 1.262,899 "step": 127900 },901 {902 "epoch": 0.8448844884488449,903 "grad_norm": 2.6924386024475098,904 "learning_rate": 0.0001810827957741188,905 "loss": 1.3436,906 "step": 128907 },908 {909 "epoch": 0.8514851485148515,910 "grad_norm": 2.324448823928833,911 "learning_rate": 0.00018062860940351916,912 "loss": 1.3375,913 "step": 129914 },915 {916 "epoch": 0.858085808580858,917 "grad_norm": 2.095839738845825,918 "learning_rate": 0.00018016961908887442,919 "loss": 1.4675,920 "step": 130921 },922 {923 "epoch": 0.8646864686468647,924 "grad_norm": 1.6898919343948364,925 "learning_rate": 0.00017970585217734844,926 "loss": 1.0081,927 "step": 131928 },929 {930 "epoch": 0.8712871287128713,931 "grad_norm": 3.500580072402954,932 "learning_rate": 0.00017923733630070002,933 "loss": 1.9837,934 "step": 132935 },936 {937 "epoch": 0.8778877887788779,938 "grad_norm": 2.213259696960449,939 "learning_rate": 0.00017876409937363677,940 "loss": 1.3967,941 "step": 133942 },943 {944 "epoch": 0.8844884488448845,945 "grad_norm": 1.764636516571045,946 "learning_rate": 0.00017828616959215184,947 "loss": 1.4183,948 "step": 134949 },950 {951 "epoch": 0.8910891089108911,952 "grad_norm": 1.8215689659118652,953 "learning_rate": 0.00017780357543184397,954 "loss": 1.5714,955 "step": 135956 },957 {958 "epoch": 0.8976897689768977,959 "grad_norm": 1.726089358329773,960 "learning_rate": 0.00017731634564622085,961 "loss": 1.4959,962 "step": 136963 },964 {965 "epoch": 0.9042904290429042,966 "grad_norm": 1.7792177200317383,967 "learning_rate": 0.00017682450926498607,968 "loss": 1.1855,969 "step": 137970 },971 {972 "epoch": 0.9108910891089109,973 "grad_norm": 1.6838842630386353,974 "learning_rate": 0.0001763280955923093,975 "loss": 1.3365,976 "step": 138977 },978 {979 "epoch": 0.9174917491749175,980 "grad_norm": 1.692114233970642,981 "learning_rate": 0.00017582713420508052,982 "loss": 1.354,983 "step": 139984 },985 {986 "epoch": 0.9240924092409241,987 "grad_norm": 1.8941043615341187,988 "learning_rate": 0.00017532165495114766,989 "loss": 1.3819,990 "step": 140991 },992 {993 "epoch": 0.9306930693069307,994 "grad_norm": 1.8271002769470215,995 "learning_rate": 0.0001748116879475383,996 "loss": 1.3082,997 "step": 141998 },999 {1000 "epoch": 0.9372937293729373,1001 "grad_norm": 2.272860288619995,1002 "learning_rate": 0.00017429726357866515,1003 "loss": 1.6585,1004 "step": 1421005 },1006 {1007 "epoch": 0.9438943894389439,1008 "grad_norm": 1.9342212677001953,1009 "learning_rate": 0.00017377841249451595,1010 "loss": 1.492,1011 "step": 1431012 },1013 {1014 "epoch": 0.9504950495049505,1015 "grad_norm": 2.028965950012207,1016 "learning_rate": 0.00017325516560882703,1017 "loss": 1.4987,1018 "step": 1441019 },1020 {1021 "epoch": 0.9570957095709571,1022 "grad_norm": 2.096977710723877,1023 "learning_rate": 0.00017272755409724168,1024 "loss": 1.489,1025 "step": 1451026 },1027 {1028 "epoch": 0.9636963696369637,1029 "grad_norm": 2.228801727294922,1030 "learning_rate": 0.00017219560939545246,1031 "loss": 1.4862,1032 "step": 1461033 },1034 {1035 "epoch": 0.9702970297029703,1036 "grad_norm": 1.7269052267074585,1037 "learning_rate": 0.0001716593631973283,1038 "loss": 1.2565,1039 "step": 1471040 },1041 {1042 "epoch": 0.976897689768977,1043 "grad_norm": 1.8832287788391113,1044 "learning_rate": 0.00017111884745302632,1045 "loss": 1.3296,1046 "step": 1481047 },1048 {1049 "epoch": 0.9834983498349835,1050 "grad_norm": 1.8937143087387085,1051 "learning_rate": 0.00017057409436708782,1052 "loss": 1.2752,1053 "step": 1491054 },1055 {1056 "epoch": 0.9900990099009901,1057 "grad_norm": 2.1348111629486084,1058 "learning_rate": 0.00017002513639651992,1059 "loss": 1.2174,1060 "step": 1501061 },1062 {1063 "epoch": 0.9966996699669967,1064 "grad_norm": 1.7981044054031372,1065 "learning_rate": 0.00016947200624886145,1066 "loss": 1.3272,1067 "step": 1511068 },1069 {1070 "epoch": 1.0,1071 "grad_norm": 3.076122999191284,1072 "learning_rate": 0.00016891473688023426,1073 "loss": 1.3359,1074 "step": 1521075 },1076 {1077 "epoch": 1.0066006600660067,1078 "grad_norm": 1.5830563306808472,1079 "learning_rate": 0.00016835336149337974,1080 "loss": 0.9539,1081 "step": 1531082 },1083 {1084 "epoch": 1.0132013201320131,1085 "grad_norm": 1.7160429954528809,1086 "learning_rate": 0.0001677879135356805,1087 "loss": 0.7741,1088 "step": 1541089 },1090 {1091 "epoch": 1.0198019801980198,1092 "grad_norm": 1.4492639303207397,1093 "learning_rate": 0.00016721842669716752,1094 "loss": 0.8243,1095 "step": 1551096 },1097 {1098 "epoch": 1.0264026402640265,1099 "grad_norm": 1.7320116758346558,1100 "learning_rate": 0.00016664493490851288,1101 "loss": 0.7563,1102 "step": 1561103 },1104 {1105 "epoch": 1.033003300330033,1106 "grad_norm": 1.8974308967590332,1107 "learning_rate": 0.00016606747233900815,1108 "loss": 0.8068,1109 "step": 1571110 },1111 {1112 "epoch": 1.0396039603960396,1113 "grad_norm": 1.688904881477356,1114 "learning_rate": 0.00016548607339452853,1115 "loss": 0.8673,1116 "step": 1581117 },1118 {1119 "epoch": 1.046204620462046,1120 "grad_norm": 1.5127521753311157,1121 "learning_rate": 0.00016490077271548285,1122 "loss": 0.6085,1123 "step": 1591124 },1125 {1126 "epoch": 1.0528052805280528,1127 "grad_norm": 1.6967127323150635,1128 "learning_rate": 0.00016431160517474985,1129 "loss": 0.6155,1130 "step": 1601131 },1132 {1133 "epoch": 1.0594059405940595,1134 "grad_norm": 2.6661343574523926,1135 "learning_rate": 0.00016371860587560008,1136 "loss": 0.8918,1137 "step": 1611138 },1139 {1140 "epoch": 1.066006600660066,1141 "grad_norm": 1.9106711149215698,1142 "learning_rate": 0.0001631218101496048,1143 "loss": 0.7613,1144 "step": 1621145 },1146 {1147 "epoch": 1.0726072607260726,1148 "grad_norm": 2.284550428390503,1149 "learning_rate": 0.0001625212535545306,1150 "loss": 0.95,1151 "step": 1631152 },1153 {1154 "epoch": 1.0792079207920793,1155 "grad_norm": 2.4170072078704834,1156 "learning_rate": 0.00016191697187222087,1157 "loss": 0.9478,1158 "step": 1641159 },1160 {1161 "epoch": 1.0858085808580857,1162 "grad_norm": 2.5793893337249756,1163 "learning_rate": 0.00016130900110646407,1164 "loss": 0.9573,1165 "step": 1651166 },1167 {1168 "epoch": 1.0924092409240924,1169 "grad_norm": 2.0098459720611572,1170 "learning_rate": 0.00016069737748084824,1171 "loss": 0.7022,1172 "step": 1661173 },1174 {1175 "epoch": 1.099009900990099,1176 "grad_norm": 1.7491934299468994,1177 "learning_rate": 0.0001600821374366031,1178 "loss": 0.727,1179 "step": 1671180 },1181 {1182 "epoch": 1.1056105610561056,1183 "grad_norm": 2.729426860809326,1184 "learning_rate": 0.00015946331763042867,1185 "loss": 0.7939,1186 "step": 1681187 },1188 {1189 "epoch": 1.1122112211221122,1190 "grad_norm": 2.475818634033203,1191 "learning_rate": 0.00015884095493231125,1192 "loss": 1.3139,1193 "step": 1691194 },1195 {1196 "epoch": 1.118811881188119,1197 "grad_norm": 2.3645308017730713,1198 "learning_rate": 0.00015821508642332657,1199 "loss": 0.6167,1200 "step": 170