NeuML/astrobert-small
115
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 45.0,6 "eval_steps": 500,7 "global_step": 158130,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.14228799089356858,14 "grad_norm": 0.6334199905395508,15 "learning_rate": 0.0001996,16 "loss": 7.48322509765625,17 "step": 50018 },19 {20 "epoch": 0.28457598178713717,21 "grad_norm": 1.0710968971252441,22 "learning_rate": 0.0003996,23 "loss": 6.2563916015625,24 "step": 100025 },26 {27 "epoch": 0.42686397268070575,28 "grad_norm": 0.6296343207359314,29 "learning_rate": 0.0005996,30 "loss": 6.11807275390625,31 "step": 150032 },33 {34 "epoch": 0.5691519635742743,35 "grad_norm": 0.9468093514442444,36 "learning_rate": 0.0007996,37 "loss": 6.00693505859375,38 "step": 200039 },40 {41 "epoch": 0.7114399544678429,42 "grad_norm": 1.3960775136947632,43 "learning_rate": 0.0009996,44 "loss": 5.5765498046875,45 "step": 250046 },47 {48 "epoch": 0.8537279453614115,49 "grad_norm": 1.1396394968032837,50 "learning_rate": 0.0009967936773115724,51 "loss": 4.27981396484375,52 "step": 300053 },54 {55 "epoch": 0.9960159362549801,56 "grad_norm": 1.1924173831939697,57 "learning_rate": 0.000993580929126775,58 "loss": 3.510773193359375,59 "step": 350060 },61 {62 "epoch": 1.1383039271485487,63 "grad_norm": 1.1864285469055176,64 "learning_rate": 0.0009903681809419777,65 "loss": 3.184400390625,66 "step": 400067 },68 {69 "epoch": 1.2805919180421172,70 "grad_norm": 0.9659832119941711,71 "learning_rate": 0.0009871554327571805,72 "loss": 2.9972861328125,73 "step": 450074 },75 {76 "epoch": 1.4228799089356858,77 "grad_norm": 1.0554136037826538,78 "learning_rate": 0.0009839426845723832,79 "loss": 2.87672265625,80 "step": 500081 },82 {83 "epoch": 1.5651678998292544,84 "grad_norm": 0.989261269569397,85 "learning_rate": 0.000980729936387586,86 "loss": 2.784982177734375,87 "step": 550088 },89 {90 "epoch": 1.707455890722823,91 "grad_norm": 1.0604394674301147,92 "learning_rate": 0.0009775171882027887,93 "loss": 2.708709716796875,94 "step": 600095 },96 {97 "epoch": 1.8497438816163916,98 "grad_norm": 0.9650368094444275,99 "learning_rate": 0.0009743044400179915,100 "loss": 2.65090625,101 "step": 6500102 },103 {104 "epoch": 1.9920318725099602,105 "grad_norm": 0.9355012774467468,106 "learning_rate": 0.0009710916918331941,107 "loss": 2.607532470703125,108 "step": 7000109 },110 {111 "epoch": 2.1343198634035287,112 "grad_norm": 0.962996780872345,113 "learning_rate": 0.0009678789436483969,114 "loss": 2.54883447265625,115 "step": 7500116 },117 {118 "epoch": 2.2766078542970973,119 "grad_norm": 1.0536885261535645,120 "learning_rate": 0.0009646661954635996,121 "loss": 2.514782958984375,122 "step": 8000123 },124 {125 "epoch": 2.418895845190666,126 "grad_norm": 1.0663453340530396,127 "learning_rate": 0.0009614534472788023,128 "loss": 2.487317138671875,129 "step": 8500130 },131 {132 "epoch": 2.5611838360842345,133 "grad_norm": 0.9626434445381165,134 "learning_rate": 0.000958240699094005,135 "loss": 2.468934814453125,136 "step": 9000137 },138 {139 "epoch": 2.703471826977803,140 "grad_norm": 0.9301921129226685,141 "learning_rate": 0.0009550279509092078,142 "loss": 2.447890380859375,143 "step": 9500144 },145 {146 "epoch": 2.8457598178713717,147 "grad_norm": 1.0070905685424805,148 "learning_rate": 0.0009518152027244105,149 "loss": 2.418142333984375,150 "step": 10000151 },152 {153 "epoch": 2.9880478087649402,154 "grad_norm": 0.9236523509025574,155 "learning_rate": 0.0009486024545396133,156 "loss": 2.40807373046875,157 "step": 10500158 },159 {160 "epoch": 3.130335799658509,161 "grad_norm": 0.9421653747558594,162 "learning_rate": 0.0009453897063548159,163 "loss": 2.368919677734375,164 "step": 11000165 },166 {167 "epoch": 3.2726237905520774,168 "grad_norm": 0.9248610138893127,169 "learning_rate": 0.0009421769581700187,170 "loss": 2.36261572265625,171 "step": 11500172 },173 {174 "epoch": 3.414911781445646,175 "grad_norm": 0.9469349384307861,176 "learning_rate": 0.0009389642099852214,177 "loss": 2.3440390625,178 "step": 12000179 },180 {181 "epoch": 3.5571997723392146,182 "grad_norm": 1.0009181499481201,183 "learning_rate": 0.0009357514618004241,184 "loss": 2.32732666015625,185 "step": 12500186 },187 {188 "epoch": 3.699487763232783,189 "grad_norm": 0.9607020020484924,190 "learning_rate": 0.0009325387136156268,191 "loss": 2.313716552734375,192 "step": 13000193 },194 {195 "epoch": 3.8417757541263517,196 "grad_norm": 0.9231477975845337,197 "learning_rate": 0.0009293259654308296,198 "loss": 2.30414453125,199 "step": 13500200 },201 {202 "epoch": 3.9840637450199203,203 "grad_norm": 0.9701341986656189,204 "learning_rate": 0.0009261132172460323,205 "loss": 2.2853681640625,206 "step": 14000207 },208 {209 "epoch": 4.126351735913489,210 "grad_norm": 0.9349779486656189,211 "learning_rate": 0.000922900469061235,212 "loss": 2.26698486328125,213 "step": 14500214 },215 {216 "epoch": 4.2686397268070575,217 "grad_norm": 0.9157093167304993,218 "learning_rate": 0.0009196877208764377,219 "loss": 2.256426513671875,220 "step": 15000221 },222 {223 "epoch": 4.410927717700626,224 "grad_norm": 0.9236478209495544,225 "learning_rate": 0.0009164749726916405,226 "loss": 2.25071337890625,227 "step": 15500228 },229 {230 "epoch": 4.553215708594195,231 "grad_norm": 0.9309902191162109,232 "learning_rate": 0.0009132622245068431,233 "loss": 2.2407861328125,234 "step": 16000235 },236 {237 "epoch": 4.695503699487763,238 "grad_norm": 0.9968065619468689,239 "learning_rate": 0.000910049476322046,240 "loss": 2.232171142578125,241 "step": 16500242 },243 {244 "epoch": 4.837791690381332,245 "grad_norm": 0.893164336681366,246 "learning_rate": 0.0009068367281372486,247 "loss": 2.223586669921875,248 "step": 17000249 },250 {251 "epoch": 4.9800796812749,252 "grad_norm": 0.9570403695106506,253 "learning_rate": 0.0009036239799524514,254 "loss": 2.2169453125,255 "step": 17500256 },257 {258 "epoch": 5.122367672168469,259 "grad_norm": 0.9152570962905884,260 "learning_rate": 0.0009004112317676541,261 "loss": 2.1965634765625,262 "step": 18000263 },264 {265 "epoch": 5.264655663062038,266 "grad_norm": 0.9538166522979736,267 "learning_rate": 0.0008971984835828568,268 "loss": 2.18993994140625,269 "step": 18500270 },271 {272 "epoch": 5.406943653955606,273 "grad_norm": 0.8849217891693115,274 "learning_rate": 0.0008939857353980595,275 "loss": 2.179218994140625,276 "step": 19000277 },278 {279 "epoch": 5.549231644849175,280 "grad_norm": 0.9768293499946594,281 "learning_rate": 0.0008907729872132623,282 "loss": 2.174207763671875,283 "step": 19500284 },285 {286 "epoch": 5.691519635742743,287 "grad_norm": 0.9277469515800476,288 "learning_rate": 0.0008875602390284649,289 "loss": 2.1700234375,290 "step": 20000291 },292 {293 "epoch": 5.833807626636312,294 "grad_norm": 0.9375103116035461,295 "learning_rate": 0.0008843474908436677,296 "loss": 2.166572998046875,297 "step": 20500298 },299 {300 "epoch": 5.9760956175298805,301 "grad_norm": 0.9594961404800415,302 "learning_rate": 0.0008811347426588704,303 "loss": 2.15810888671875,304 "step": 21000305 },306 {307 "epoch": 6.118383608423449,308 "grad_norm": 0.8761999607086182,309 "learning_rate": 0.0008779219944740731,310 "loss": 2.13630029296875,311 "step": 21500312 },313 {314 "epoch": 6.260671599317018,315 "grad_norm": 0.8995809555053711,316 "learning_rate": 0.0008747092462892759,317 "loss": 2.13485693359375,318 "step": 22000319 },320 {321 "epoch": 6.402959590210586,322 "grad_norm": 0.9557548761367798,323 "learning_rate": 0.0008714964981044786,324 "loss": 2.12809130859375,325 "step": 22500326 },327 {328 "epoch": 6.545247581104155,329 "grad_norm": 0.8720033168792725,330 "learning_rate": 0.0008682837499196813,331 "loss": 2.12343359375,332 "step": 23000333 },334 {335 "epoch": 6.687535571997723,336 "grad_norm": 0.9293733239173889,337 "learning_rate": 0.0008650710017348841,338 "loss": 2.121170166015625,339 "step": 23500340 },341 {342 "epoch": 6.829823562891292,343 "grad_norm": 0.9301312565803528,344 "learning_rate": 0.0008618582535500867,345 "loss": 2.115037841796875,346 "step": 24000347 },348 {349 "epoch": 6.972111553784861,350 "grad_norm": 0.9322926998138428,351 "learning_rate": 0.0008586455053652895,352 "loss": 2.11170068359375,353 "step": 24500354 },355 {356 "epoch": 7.114399544678429,357 "grad_norm": 0.9266783595085144,358 "learning_rate": 0.0008554327571804922,359 "loss": 2.0983544921875,360 "step": 25000361 },362 {363 "epoch": 7.256687535571998,364 "grad_norm": 0.9406460523605347,365 "learning_rate": 0.0008522200089956949,366 "loss": 2.08985888671875,367 "step": 25500368 },369 {370 "epoch": 7.398975526465566,371 "grad_norm": 0.923007607460022,372 "learning_rate": 0.0008490072608108977,373 "loss": 2.093153564453125,374 "step": 26000375 },376 {377 "epoch": 7.541263517359135,378 "grad_norm": 0.9110872149467468,379 "learning_rate": 0.0008457945126261004,380 "loss": 2.080158447265625,381 "step": 26500382 },383 {384 "epoch": 7.6835515082527035,385 "grad_norm": 0.9028299450874329,386 "learning_rate": 0.000842581764441303,387 "loss": 2.0810068359375,388 "step": 27000389 },390 {391 "epoch": 7.825839499146272,392 "grad_norm": 0.9071224927902222,393 "learning_rate": 0.0008393690162565059,394 "loss": 2.08135595703125,395 "step": 27500396 },397 {398 "epoch": 7.968127490039841,399 "grad_norm": 0.8702104687690735,400 "learning_rate": 0.0008361562680717085,401 "loss": 2.073836181640625,402 "step": 28000403 },404 {405 "epoch": 8.110415480933408,406 "grad_norm": 0.9875132441520691,407 "learning_rate": 0.0008329435198869114,408 "loss": 2.06207470703125,409 "step": 28500410 },411 {412 "epoch": 8.252703471826978,413 "grad_norm": 0.8733546733856201,414 "learning_rate": 0.000829730771702114,415 "loss": 2.0590986328125,416 "step": 29000417 },418 {419 "epoch": 8.394991462720546,420 "grad_norm": 0.8623999357223511,421 "learning_rate": 0.0008265180235173167,422 "loss": 2.05387744140625,423 "step": 29500424 },425 {426 "epoch": 8.537279453614115,427 "grad_norm": 0.9093107581138611,428 "learning_rate": 0.0008233052753325195,429 "loss": 2.049600830078125,430 "step": 30000431 },432 {433 "epoch": 8.679567444507683,434 "grad_norm": 0.8935321569442749,435 "learning_rate": 0.0008200925271477222,436 "loss": 2.04915966796875,437 "step": 30500438 },439 {440 "epoch": 8.821855435401252,441 "grad_norm": 0.9517919421195984,442 "learning_rate": 0.0008168797789629248,443 "loss": 2.044854248046875,444 "step": 31000445 },446 {447 "epoch": 8.964143426294822,448 "grad_norm": 0.924595057964325,449 "learning_rate": 0.0008136670307781277,450 "loss": 2.0417685546875,451 "step": 31500452 },453 {454 "epoch": 9.10643141718839,455 "grad_norm": 0.8856834173202515,456 "learning_rate": 0.0008104542825933303,457 "loss": 2.0327542724609375,458 "step": 32000459 },460 {461 "epoch": 9.248719408081957,462 "grad_norm": 0.851649820804596,463 "learning_rate": 0.000807241534408533,464 "loss": 2.026624267578125,465 "step": 32500466 },467 {468 "epoch": 9.391007398975526,469 "grad_norm": 0.8763725161552429,470 "learning_rate": 0.0008040287862237358,471 "loss": 2.023781982421875,472 "step": 33000473 },474 {475 "epoch": 9.533295389869096,476 "grad_norm": 0.9054471850395203,477 "learning_rate": 0.0008008160380389385,478 "loss": 2.0217911376953124,479 "step": 33500480 },481 {482 "epoch": 9.675583380762664,483 "grad_norm": 0.9936219453811646,484 "learning_rate": 0.0007976032898541413,485 "loss": 2.0223516845703124,486 "step": 34000487 },488 {489 "epoch": 9.817871371656231,490 "grad_norm": 0.8445395231246948,491 "learning_rate": 0.000794390541669344,492 "loss": 2.0150150146484376,493 "step": 34500494 },495 {496 "epoch": 9.9601593625498,497 "grad_norm": 0.8997848629951477,498 "learning_rate": 0.0007911777934845466,499 "loss": 2.0144486083984376,500 "step": 35000501 },502 {503 "epoch": 10.102447353443369,504 "grad_norm": 0.9022369384765625,505 "learning_rate": 0.0007879650452997495,506 "loss": 2.0057178955078125,507 "step": 35500508 },509 {510 "epoch": 10.244735344336938,511 "grad_norm": 0.8679398894309998,512 "learning_rate": 0.0007847522971149521,513 "loss": 1.996031005859375,514 "step": 36000515 },516 {517 "epoch": 10.387023335230506,518 "grad_norm": 0.8584362864494324,519 "learning_rate": 0.0007815395489301548,520 "loss": 1.9979686279296875,521 "step": 36500522 },523 {524 "epoch": 10.529311326124075,525 "grad_norm": 0.919893741607666,526 "learning_rate": 0.0007783268007453576,527 "loss": 1.9953302001953126,528 "step": 37000529 },530 {531 "epoch": 10.671599317017645,532 "grad_norm": 0.8829244375228882,533 "learning_rate": 0.0007751140525605603,534 "loss": 1.9904837646484375,535 "step": 37500536 },537 {538 "epoch": 10.813887307911212,539 "grad_norm": 0.8700593113899231,540 "learning_rate": 0.000771901304375763,541 "loss": 1.9878653564453126,542 "step": 38000543 },544 {545 "epoch": 10.95617529880478,546 "grad_norm": 0.8948189616203308,547 "learning_rate": 0.0007686885561909658,548 "loss": 1.9861346435546876,549 "step": 38500550 },551 {552 "epoch": 11.09846328969835,553 "grad_norm": 0.8743970990180969,554 "learning_rate": 0.0007654758080061684,555 "loss": 1.9781466064453126,556 "step": 39000557 },558 {559 "epoch": 11.240751280591917,560 "grad_norm": 0.8927684426307678,561 "learning_rate": 0.0007622630598213713,562 "loss": 1.9800474853515626,563 "step": 39500564 },565 {566 "epoch": 11.383039271485487,567 "grad_norm": 0.8602464199066162,568 "learning_rate": 0.0007590503116365739,569 "loss": 1.969161865234375,570 "step": 40000571 },572 {573 "epoch": 11.525327262379054,574 "grad_norm": 0.9330834746360779,575 "learning_rate": 0.0007558375634517766,576 "loss": 1.9744647216796876,577 "step": 40500578 },579 {580 "epoch": 11.667615253272624,581 "grad_norm": 0.8487387895584106,582 "learning_rate": 0.0007526248152669794,583 "loss": 1.9631361083984376,584 "step": 41000585 },586 {587 "epoch": 11.809903244166193,588 "grad_norm": 0.8121556639671326,589 "learning_rate": 0.0007494120670821821,590 "loss": 1.9670977783203125,591 "step": 41500592 },593 {594 "epoch": 11.952191235059761,595 "grad_norm": 0.8926526308059692,596 "learning_rate": 0.0007461993188973848,597 "loss": 1.962933349609375,598 "step": 42000599 },600 {601 "epoch": 12.094479225953329,602 "grad_norm": 0.8671021461486816,603 "learning_rate": 0.0007429865707125876,604 "loss": 1.9563138427734375,605 "step": 42500606 },607 {608 "epoch": 12.236767216846898,609 "grad_norm": 0.8316618800163269,610 "learning_rate": 0.0007397738225277902,611 "loss": 1.9496668701171875,612 "step": 43000613 },614 {615 "epoch": 12.379055207740466,616 "grad_norm": 0.8220306038856506,617 "learning_rate": 0.000736561074342993,618 "loss": 1.9505400390625,619 "step": 43500620 },621 {622 "epoch": 12.521343198634035,623 "grad_norm": 0.882720410823822,624 "learning_rate": 0.0007333483261581957,625 "loss": 1.9462012939453126,626 "step": 44000627 },628 {629 "epoch": 12.663631189527603,630 "grad_norm": 0.8841201663017273,631 "learning_rate": 0.0007301355779733984,632 "loss": 1.94572216796875,633 "step": 44500634 },635 {636 "epoch": 12.805919180421172,637 "grad_norm": 0.8917742967605591,638 "learning_rate": 0.0007269228297886012,639 "loss": 1.9462489013671875,640 "step": 45000641 },642 {643 "epoch": 12.948207171314742,644 "grad_norm": 0.8612926006317139,645 "learning_rate": 0.0007237100816038039,646 "loss": 1.9503948974609375,647 "step": 45500648 },649 {650 "epoch": 13.09049516220831,651 "grad_norm": 0.9925726652145386,652 "learning_rate": 0.0007204973334190066,653 "loss": 1.9372962646484375,654 "step": 46000655 },656 {657 "epoch": 13.232783153101877,658 "grad_norm": 0.8212889432907104,659 "learning_rate": 0.0007172845852342094,660 "loss": 1.93100146484375,661 "step": 46500662 },663 {664 "epoch": 13.375071143995447,665 "grad_norm": 0.8393476605415344,666 "learning_rate": 0.000714071837049412,667 "loss": 1.9355791015625,668 "step": 47000669 },670 {671 "epoch": 13.517359134889016,672 "grad_norm": 0.876530647277832,673 "learning_rate": 0.0007108590888646148,674 "loss": 1.929031982421875,675 "step": 47500676 },677 {678 "epoch": 13.659647125782584,679 "grad_norm": 0.9148900508880615,680 "learning_rate": 0.0007076463406798175,681 "loss": 1.9321796875,682 "step": 48000683 },684 {685 "epoch": 13.801935116676152,686 "grad_norm": 0.8540393114089966,687 "learning_rate": 0.0007044335924950202,688 "loss": 1.9232811279296875,689 "step": 48500690 },691 {692 "epoch": 13.944223107569721,693 "grad_norm": 0.874427080154419,694 "learning_rate": 0.0007012208443102229,695 "loss": 1.920580078125,696 "step": 49000697 },698 {699 "epoch": 14.086511098463289,700 "grad_norm": 0.9040400385856628,701 "learning_rate": 0.0006980080961254257,702 "loss": 1.912923095703125,703 "step": 49500704 },705 {706 "epoch": 14.228799089356858,707 "grad_norm": 0.9279779195785522,708 "learning_rate": 0.0006947953479406284,709 "loss": 1.9119281005859374,710 "step": 50000711 },712 {713 "epoch": 14.371087080250426,714 "grad_norm": 0.8857927322387695,715 "learning_rate": 0.0006915825997558312,716 "loss": 1.907870361328125,717 "step": 50500718 },719 {720 "epoch": 14.513375071143995,721 "grad_norm": 0.8245786428451538,722 "learning_rate": 0.0006883698515710338,723 "loss": 1.9115518798828126,724 "step": 51000725 },726 {727 "epoch": 14.655663062037565,728 "grad_norm": 0.8284105658531189,729 "learning_rate": 0.0006851571033862366,730 "loss": 1.91026611328125,731 "step": 51500732 },733 {734 "epoch": 14.797951052931133,735 "grad_norm": 0.8700644969940186,736 "learning_rate": 0.0006819443552014393,737 "loss": 1.9089705810546875,738 "step": 52000739 },740 {741 "epoch": 14.9402390438247,742 "grad_norm": 0.8545106649398804,743 "learning_rate": 0.000678731607016642,744 "loss": 1.9038966064453124,745 "step": 52500746 },747 {748 "epoch": 15.08252703471827,749 "grad_norm": 0.8972774744033813,750 "learning_rate": 0.0006755188588318448,751 "loss": 1.9033292236328125,752 "step": 53000753 },754 {755 "epoch": 15.224815025611838,756 "grad_norm": 0.8351185321807861,757 "learning_rate": 0.0006723061106470475,758 "loss": 1.8937845458984375,759 "step": 53500760 },761 {762 "epoch": 15.367103016505407,763 "grad_norm": 0.8219364285469055,764 "learning_rate": 0.0006690933624622503,765 "loss": 1.8965240478515626,766 "step": 54000767 },768 {769 "epoch": 15.509391007398975,770 "grad_norm": 0.8361454606056213,771 "learning_rate": 0.0006658806142774529,772 "loss": 1.89189599609375,773 "step": 54500774 },775 {776 "epoch": 15.651678998292544,777 "grad_norm": 0.8974409699440002,778 "learning_rate": 0.0006626678660926557,779 "loss": 1.89478369140625,780 "step": 55000781 },782 {783 "epoch": 15.793966989186112,784 "grad_norm": 0.880893886089325,785 "learning_rate": 0.0006594551179078584,786 "loss": 1.8854979248046875,787 "step": 55500788 },789 {790 "epoch": 15.936254980079681,791 "grad_norm": 0.8354572057723999,792 "learning_rate": 0.0006562423697230612,793 "loss": 1.884849853515625,794 "step": 56000795 },796 {797 "epoch": 16.07854297097325,798 "grad_norm": 0.8348339200019836,799 "learning_rate": 0.0006530296215382638,800 "loss": 1.8829962158203124,801 "step": 56500802 },803 {804 "epoch": 16.220830961866817,805 "grad_norm": 0.828911304473877,806 "learning_rate": 0.0006498168733534666,807 "loss": 1.8815081787109376,808 "step": 57000809 },810 {811 "epoch": 16.363118952760388,812 "grad_norm": 0.8096295595169067,813 "learning_rate": 0.0006466041251686693,814 "loss": 1.87695361328125,815 "step": 57500816 },817 {818 "epoch": 16.505406943653956,819 "grad_norm": 0.8548514246940613,820 "learning_rate": 0.0006433913769838721,821 "loss": 1.8754146728515626,822 "step": 58000823 },824 {825 "epoch": 16.647694934547523,826 "grad_norm": 0.8417186737060547,827 "learning_rate": 0.0006401786287990747,828 "loss": 1.8718682861328124,829 "step": 58500830 },831 {832 "epoch": 16.78998292544109,833 "grad_norm": 0.8816096186637878,834 "learning_rate": 0.0006369658806142775,835 "loss": 1.876389404296875,836 "step": 59000837 },838 {839 "epoch": 16.932270916334662,840 "grad_norm": 0.8592162728309631,841 "learning_rate": 0.0006337531324294802,842 "loss": 1.8765806884765626,843 "step": 59500844 },845 {846 "epoch": 17.07455890722823,847 "grad_norm": 0.8253895044326782,848 "learning_rate": 0.0006305403842446829,849 "loss": 1.862882568359375,850 "step": 60000851 },852 {853 "epoch": 17.216846898121798,854 "grad_norm": 0.848976731300354,855 "learning_rate": 0.0006273276360598856,856 "loss": 1.860851806640625,857 "step": 60500858 },859 {860 "epoch": 17.359134889015365,861 "grad_norm": 0.8160614967346191,862 "learning_rate": 0.0006241148878750884,863 "loss": 1.8602349853515625,864 "step": 61000865 },866 {867 "epoch": 17.501422879908937,868 "grad_norm": 0.8240634799003601,869 "learning_rate": 0.0006209021396902911,870 "loss": 1.855229736328125,871 "step": 61500872 },873 {874 "epoch": 17.643710870802504,875 "grad_norm": 0.82338947057724,876 "learning_rate": 0.0006176893915054939,877 "loss": 1.8570489501953125,878 "step": 62000879 },880 {881 "epoch": 17.785998861696072,882 "grad_norm": 0.7808911204338074,883 "learning_rate": 0.0006144766433206965,884 "loss": 1.8573056640625,885 "step": 62500886 },887 {888 "epoch": 17.92828685258964,889 "grad_norm": 0.8358622789382935,890 "learning_rate": 0.0006112638951358993,891 "loss": 1.860720703125,892 "step": 63000893 },894 {895 "epoch": 18.07057484348321,896 "grad_norm": 0.8733948469161987,897 "learning_rate": 0.000608051146951102,898 "loss": 1.8501636962890624,899 "step": 63500900 },901 {902 "epoch": 18.21286283437678,903 "grad_norm": 0.8293560743331909,904 "learning_rate": 0.0006048383987663047,905 "loss": 1.8436048583984375,906 "step": 64000907 },908 {909 "epoch": 18.355150825270346,910 "grad_norm": 0.8083025217056274,911 "learning_rate": 0.0006016256505815074,912 "loss": 1.845773681640625,913 "step": 64500914 },915 {916 "epoch": 18.497438816163914,917 "grad_norm": 0.8062528967857361,918 "learning_rate": 0.0005984129023967102,919 "loss": 1.849447998046875,920 "step": 65000921 },922 {923 "epoch": 18.639726807057485,924 "grad_norm": 0.797907829284668,925 "learning_rate": 0.0005952001542119128,926 "loss": 1.8439942626953125,927 "step": 65500928 },929 {930 "epoch": 18.782014797951053,931 "grad_norm": 0.8070884346961975,932 "learning_rate": 0.0005919874060271157,933 "loss": 1.8405416259765626,934 "step": 66000935 },936 {937 "epoch": 18.92430278884462,938 "grad_norm": 0.8484400510787964,939 "learning_rate": 0.0005887746578423183,940 "loss": 1.8404969482421876,941 "step": 66500942 },943 {944 "epoch": 19.06659077973819,945 "grad_norm": 0.8148825168609619,946 "learning_rate": 0.0005855619096575211,947 "loss": 1.8372745361328124,948 "step": 67000949 },950 {951 "epoch": 19.20887877063176,952 "grad_norm": 0.7816134691238403,953 "learning_rate": 0.0005823491614727238,954 "loss": 1.8346165771484375,955 "step": 67500956 },957 {958 "epoch": 19.351166761525327,959 "grad_norm": 0.8447558283805847,960 "learning_rate": 0.0005791364132879265,961 "loss": 1.8296884765625,962 "step": 68000963 },964 {965 "epoch": 19.493454752418895,966 "grad_norm": 0.8260893225669861,967 "learning_rate": 0.0005759236651031293,968 "loss": 1.830061279296875,969 "step": 68500970 },971 {972 "epoch": 19.635742743312463,973 "grad_norm": 0.7893286347389221,974 "learning_rate": 0.000572710916918332,975 "loss": 1.829275634765625,976 "step": 69000977 },978 {979 "epoch": 19.778030734206034,980 "grad_norm": 0.8122330904006958,981 "learning_rate": 0.0005694981687335346,982 "loss": 1.829096923828125,983 "step": 69500984 },985 {986 "epoch": 19.9203187250996,987 "grad_norm": 0.825334370136261,988 "learning_rate": 0.0005662854205487375,989 "loss": 1.82976708984375,990 "step": 70000991 },992 {993 "epoch": 20.06260671599317,994 "grad_norm": 0.8224254846572876,995 "learning_rate": 0.0005630726723639401,996 "loss": 1.8201314697265625,997 "step": 70500998 },999 {1000 "epoch": 20.204894706886737,1001 "grad_norm": 0.8266887068748474,1002 "learning_rate": 0.0005598599241791428,1003 "loss": 1.821681884765625,1004 "step": 710001005 },1006 {1007 "epoch": 20.34718269778031,1008 "grad_norm": 0.8528222441673279,1009 "learning_rate": 0.0005566471759943456,1010 "loss": 1.813265869140625,1011 "step": 715001012 },1013 {1014 "epoch": 20.489470688673876,1015 "grad_norm": 0.8061295747756958,1016 "learning_rate": 0.0005534344278095483,1017 "loss": 1.81572412109375,1018 "step": 720001019 },1020 {1021 "epoch": 20.631758679567444,1022 "grad_norm": 0.8042652010917664,1023 "learning_rate": 0.000550221679624751,1024 "loss": 1.8168936767578125,1025 "step": 725001026 },1027 {1028 "epoch": 20.77404667046101,1029 "grad_norm": 0.7869358658790588,1030 "learning_rate": 0.0005470089314399538,1031 "loss": 1.814272705078125,1032 "step": 730001033 },1034 {1035 "epoch": 20.916334661354583,1036 "grad_norm": 0.8155378699302673,1037 "learning_rate": 0.0005437961832551564,1038 "loss": 1.81436279296875,1039 "step": 735001040 },1041 {1042 "epoch": 21.05862265224815,1043 "grad_norm": 0.7901885509490967,1044 "learning_rate": 0.0005405834350703593,1045 "loss": 1.811807861328125,1046 "step": 740001047 },1048 {1049 "epoch": 21.200910643141718,1050 "grad_norm": 0.7788444757461548,1051 "learning_rate": 0.0005373706868855619,1052 "loss": 1.8064168701171874,1053 "step": 745001054 },1055 {1056 "epoch": 21.343198634035286,1057 "grad_norm": 0.8150326609611511,1058 "learning_rate": 0.0005341579387007646,1059 "loss": 1.7998411865234376,1060 "step": 750001061 },1062 {1063 "epoch": 21.485486624928857,1064 "grad_norm": 0.790658175945282,1065 "learning_rate": 0.0005309451905159674,1066 "loss": 1.8009261474609375,1067 "step": 755001068 },1069 {1070 "epoch": 21.627774615822425,1071 "grad_norm": 0.8362455368041992,1072 "learning_rate": 0.0005277324423311701,1073 "loss": 1.7981884765625,1074 "step": 760001075 },1076 {1077 "epoch": 21.770062606715992,1078 "grad_norm": 0.8072263598442078,1079 "learning_rate": 0.0005245196941463727,1080 "loss": 1.802354248046875,1081 "step": 765001082 },1083 {1084 "epoch": 21.91235059760956,1085 "grad_norm": 0.8608630895614624,1086 "learning_rate": 0.0005213069459615756,1087 "loss": 1.7926978759765626,1088 "step": 770001089 },1090 {1091 "epoch": 22.05463858850313,1092 "grad_norm": 0.8350149989128113,1093 "learning_rate": 0.0005180941977767782,1094 "loss": 1.7972198486328126,1095 "step": 775001096 },1097 {1098 "epoch": 22.1969265793967,1099 "grad_norm": 0.8605798482894897,1100 "learning_rate": 0.0005148814495919811,1101 "loss": 1.7945047607421876,1102 "step": 780001103 },1104 {1105 "epoch": 22.339214570290267,1106 "grad_norm": 0.8054996728897095,1107 "learning_rate": 0.0005116687014071837,1108 "loss": 1.7889041748046874,1109 "step": 785001110 },1111 {1112 "epoch": 22.481502561183834,1113 "grad_norm": 0.8018432855606079,1114 "learning_rate": 0.0005084559532223864,1115 "loss": 1.7955711669921874,1116 "step": 790001117 },1118 {1119 "epoch": 22.623790552077406,1120 "grad_norm": 0.8167839050292969,1121 "learning_rate": 0.0005052432050375892,1122 "loss": 1.789824462890625,1123 "step": 795001124 },1125 {1126 "epoch": 22.766078542970973,1127 "grad_norm": 0.8051914572715759,1128 "learning_rate": 0.0005020304568527919,1129 "loss": 1.786076904296875,1130 "step": 800001131 },1132 {1133 "epoch": 22.90836653386454,1134 "grad_norm": 0.8690944314002991,1135 "learning_rate": 0.0004988177086679945,1136 "loss": 1.787760009765625,1137 "step": 805001138 },1139 {1140 "epoch": 23.05065452475811,1141 "grad_norm": 0.7857160568237305,1142 "learning_rate": 0.0004956049604831974,1143 "loss": 1.78182080078125,1144 "step": 810001145 },1146 {1147 "epoch": 23.19294251565168,1148 "grad_norm": 0.792676568031311,1149 "learning_rate": 0.0004923922122984001,1150 "loss": 1.7775865478515624,1151 "step": 815001152 },1153 {1154 "epoch": 23.335230506545248,1155 "grad_norm": 0.8044800162315369,1156 "learning_rate": 0.0004891794641136028,1157 "loss": 1.776404052734375,1158 "step": 820001159 },1160 {1161 "epoch": 23.477518497438815,1162 "grad_norm": 0.8092531561851501,1163 "learning_rate": 0.00048596671592880555,1164 "loss": 1.7750899658203125,1165 "step": 825001166 },1167 {1168 "epoch": 23.619806488332383,1169 "grad_norm": 0.8156195878982544,1170 "learning_rate": 0.00048275396774400824,1171 "loss": 1.7721761474609374,1172 "step": 830001173 },1174 {1175 "epoch": 23.762094479225954,1176 "grad_norm": 0.8054773211479187,1177 "learning_rate": 0.000479541219559211,1178 "loss": 1.7737218017578125,1179 "step": 835001180 },1181 {1182 "epoch": 23.904382470119522,1183 "grad_norm": 0.8544410467147827,1184 "learning_rate": 0.0004763284713744137,1185 "loss": 1.7725914306640624,1186 "step": 840001187 },1188 {1189 "epoch": 24.04667046101309,1190 "grad_norm": 0.8268908262252808,1191 "learning_rate": 0.00047311572318961645,1192 "loss": 1.7661802978515626,1193 "step": 845001194 },1195 {1196 "epoch": 24.188958451906657,1197 "grad_norm": 0.8119874000549316,1198 "learning_rate": 0.00046990297500481914,1199 "loss": 1.7655081787109375,1200 "step": 85000