impossibleexchange/dipft
05
1{2 "best_metric": null,3 "best_model_checkpoint": null,4 "epoch": 1.0,5 "eval_steps": 4064,6 "global_step": 16256,7 "is_hyper_param_search": false,8 "is_local_process_zero": true,9 "is_world_process_zero": true,10 "log_history": [11 {12 "epoch": 6.151574803149606e-05,13 "grad_norm": 5.482029914855957,14 "learning_rate": 2e-05,15 "loss": 13.6418,16 "step": 117 },18 {19 "epoch": 6.151574803149606e-05,20 "eval_loss": 14.886382102966309,21 "eval_runtime": 1039.1143,22 "eval_samples_per_second": 3.69,23 "eval_steps_per_second": 3.69,24 "step": 125 },26 {27 "epoch": 0.00012303149606299212,28 "grad_norm": 5.984772205352783,29 "learning_rate": 4e-05,30 "loss": 15.0598,31 "step": 232 },33 {34 "epoch": 0.00018454724409448819,35 "grad_norm": 5.734090805053711,36 "learning_rate": 6e-05,37 "loss": 16.4218,38 "step": 339 },40 {41 "epoch": 0.00024606299212598425,42 "grad_norm": 5.444360256195068,43 "learning_rate": 8e-05,44 "loss": 12.9511,45 "step": 446 },47 {48 "epoch": 0.0003075787401574803,49 "grad_norm": 5.957120418548584,50 "learning_rate": 0.0001,51 "loss": 15.2019,52 "step": 553 },54 {55 "epoch": 0.00036909448818897637,56 "grad_norm": 6.127737045288086,57 "learning_rate": 0.00012,58 "loss": 16.6303,59 "step": 660 },61 {62 "epoch": 0.00043061023622047243,63 "grad_norm": 6.821583271026611,64 "learning_rate": 0.00014,65 "loss": 13.9561,66 "step": 767 },68 {69 "epoch": 0.0004921259842519685,70 "grad_norm": 6.55456018447876,71 "learning_rate": 0.00016,72 "loss": 10.5351,73 "step": 874 },75 {76 "epoch": 0.0005536417322834646,77 "grad_norm": 8.733551025390625,78 "learning_rate": 0.00018,79 "loss": 12.0014,80 "step": 981 },82 {83 "epoch": 0.0006151574803149606,84 "grad_norm": 5.629597187042236,85 "learning_rate": 0.0002,86 "loss": 10.9767,87 "step": 1088 },89 {90 "epoch": 0.0006766732283464567,91 "grad_norm": 9.613995552062988,92 "learning_rate": 0.00019999999813027883,93 "loss": 9.8536,94 "step": 1195 },96 {97 "epoch": 0.0007381889763779527,98 "grad_norm": 5.925160884857178,99 "learning_rate": 0.0001999999925211153,100 "loss": 8.3939,101 "step": 12102 },103 {104 "epoch": 0.0007997047244094488,105 "grad_norm": 5.521446704864502,106 "learning_rate": 0.0001999999831725097,107 "loss": 8.6053,108 "step": 13109 },110 {111 "epoch": 0.0008612204724409449,112 "grad_norm": 4.685490131378174,113 "learning_rate": 0.00019999997008446232,114 "loss": 7.699,115 "step": 14116 },117 {118 "epoch": 0.0009227362204724409,119 "grad_norm": 3.1345713138580322,120 "learning_rate": 0.0001999999532569737,121 "loss": 7.0154,122 "step": 15123 },124 {125 "epoch": 0.000984251968503937,126 "grad_norm": 2.5838406085968018,127 "learning_rate": 0.0001999999326900444,128 "loss": 4.7423,129 "step": 16130 },131 {132 "epoch": 0.0010457677165354332,133 "grad_norm": 2.0108108520507812,134 "learning_rate": 0.0001999999083836753,135 "loss": 5.5067,136 "step": 17137 },138 {139 "epoch": 0.0011072834645669291,140 "grad_norm": 2.3351149559020996,141 "learning_rate": 0.0001999998803378672,142 "loss": 5.7271,143 "step": 18144 },145 {146 "epoch": 0.0011687992125984253,147 "grad_norm": 2.063504219055176,148 "learning_rate": 0.0001999998485526212,149 "loss": 5.4698,150 "step": 19151 },152 {153 "epoch": 0.0012303149606299212,154 "grad_norm": 2.36728572845459,155 "learning_rate": 0.00019999981302793845,156 "loss": 5.7157,157 "step": 20158 },159 {160 "epoch": 0.0012918307086614174,161 "grad_norm": 1.942466139793396,162 "learning_rate": 0.00019999977376382032,163 "loss": 5.8405,164 "step": 21165 },166 {167 "epoch": 0.0013533464566929134,168 "grad_norm": 2.3639914989471436,169 "learning_rate": 0.0001999997307602683,170 "loss": 4.7815,171 "step": 22172 },173 {174 "epoch": 0.0014148622047244095,175 "grad_norm": 2.7135772705078125,176 "learning_rate": 0.00019999968401728392,177 "loss": 5.0877,178 "step": 23179 },180 {181 "epoch": 0.0014763779527559055,182 "grad_norm": 4.7281622886657715,183 "learning_rate": 0.00019999963353486898,184 "loss": 6.4028,185 "step": 24186 },187 {188 "epoch": 0.0015378937007874017,189 "grad_norm": 2.477782964706421,190 "learning_rate": 0.0001999995793130254,191 "loss": 5.158,192 "step": 25193 },194 {195 "epoch": 0.0015994094488188976,196 "grad_norm": 1.902949333190918,197 "learning_rate": 0.0001999995213517551,198 "loss": 4.7101,199 "step": 26200 },201 {202 "epoch": 0.0016609251968503938,203 "grad_norm": 2.387880563735962,204 "learning_rate": 0.00019999945965106037,205 "loss": 5.9719,206 "step": 27207 },208 {209 "epoch": 0.0017224409448818897,210 "grad_norm": 1.3084725141525269,211 "learning_rate": 0.00019999939421094345,212 "loss": 5.4829,213 "step": 28214 },215 {216 "epoch": 0.001783956692913386,217 "grad_norm": 1.1015849113464355,218 "learning_rate": 0.00019999932503140676,219 "loss": 5.2135,220 "step": 29221 },222 {223 "epoch": 0.0018454724409448819,224 "grad_norm": 0.9842422008514404,225 "learning_rate": 0.00019999925211245295,226 "loss": 5.1587,227 "step": 30228 },229 {230 "epoch": 0.001906988188976378,231 "grad_norm": 1.2972447872161865,232 "learning_rate": 0.00019999917545408474,233 "loss": 5.0864,234 "step": 31235 },236 {237 "epoch": 0.001968503937007874,238 "grad_norm": 1.1920126676559448,239 "learning_rate": 0.00019999909505630497,240 "loss": 4.7727,241 "step": 32242 },243 {244 "epoch": 0.00203001968503937,245 "grad_norm": 1.1871001720428467,246 "learning_rate": 0.00019999901091911662,247 "loss": 4.4437,248 "step": 33249 },250 {251 "epoch": 0.0020915354330708663,252 "grad_norm": 1.1514976024627686,253 "learning_rate": 0.00019999892304252293,254 "loss": 5.1533,255 "step": 34256 },257 {258 "epoch": 0.0021530511811023623,259 "grad_norm": 0.9348739385604858,260 "learning_rate": 0.00019999883142652708,261 "loss": 4.5989,262 "step": 35263 },264 {265 "epoch": 0.0022145669291338582,266 "grad_norm": 1.83991277217865,267 "learning_rate": 0.00019999873607113257,268 "loss": 4.7957,269 "step": 36270 },271 {272 "epoch": 0.002276082677165354,273 "grad_norm": 0.9639754891395569,274 "learning_rate": 0.00019999863697634293,275 "loss": 5.1886,276 "step": 37277 },278 {279 "epoch": 0.0023375984251968506,280 "grad_norm": 0.8030492067337036,281 "learning_rate": 0.00019999853414216188,282 "loss": 4.9723,283 "step": 38284 },285 {286 "epoch": 0.0023991141732283465,287 "grad_norm": 0.7356982231140137,288 "learning_rate": 0.00019999842756859325,289 "loss": 4.5846,290 "step": 39291 },292 {293 "epoch": 0.0024606299212598425,294 "grad_norm": 0.6708055734634399,295 "learning_rate": 0.00019999831725564103,296 "loss": 4.3259,297 "step": 40298 },299 {300 "epoch": 0.0025221456692913384,301 "grad_norm": 0.8197684288024902,302 "learning_rate": 0.00019999820320330938,303 "loss": 4.4742,304 "step": 41305 },306 {307 "epoch": 0.002583661417322835,308 "grad_norm": 0.8065810203552246,309 "learning_rate": 0.00019999808541160252,310 "loss": 4.4442,311 "step": 42312 },313 {314 "epoch": 0.0026451771653543308,315 "grad_norm": 1.2500178813934326,316 "learning_rate": 0.00019999796388052486,317 "loss": 4.7131,318 "step": 43319 },320 {321 "epoch": 0.0027066929133858267,322 "grad_norm": 0.6776506304740906,323 "learning_rate": 0.00019999783861008097,324 "loss": 4.4275,325 "step": 44326 },327 {328 "epoch": 0.0027682086614173227,329 "grad_norm": 0.8938648700714111,330 "learning_rate": 0.0001999977096002755,331 "loss": 2.8587,332 "step": 45333 },334 {335 "epoch": 0.002829724409448819,336 "grad_norm": 0.8551689982414246,337 "learning_rate": 0.0001999975768511133,338 "loss": 5.0646,339 "step": 46340 },341 {342 "epoch": 0.002891240157480315,343 "grad_norm": 1.5305277109146118,344 "learning_rate": 0.00019999744036259935,345 "loss": 6.1155,346 "step": 47347 },348 {349 "epoch": 0.002952755905511811,350 "grad_norm": 1.9128812551498413,351 "learning_rate": 0.0001999973001347387,352 "loss": 5.0263,353 "step": 48354 },355 {356 "epoch": 0.003014271653543307,357 "grad_norm": 1.0714994668960571,358 "learning_rate": 0.00019999715616753666,359 "loss": 4.6472,360 "step": 49361 },362 {363 "epoch": 0.0030757874015748033,364 "grad_norm": 0.7470815777778625,365 "learning_rate": 0.00019999700846099855,366 "loss": 4.1215,367 "step": 50368 },369 {370 "epoch": 0.0031373031496062993,371 "grad_norm": 1.3089985847473145,372 "learning_rate": 0.00019999685701512992,373 "loss": 4.2695,374 "step": 51375 },376 {377 "epoch": 0.0031988188976377952,378 "grad_norm": 1.1406008005142212,379 "learning_rate": 0.00019999670182993645,380 "loss": 4.2568,381 "step": 52382 },383 {384 "epoch": 0.003260334645669291,385 "grad_norm": 1.4149187803268433,386 "learning_rate": 0.00019999654290542393,387 "loss": 4.1756,388 "step": 53389 },390 {391 "epoch": 0.0033218503937007876,392 "grad_norm": 0.8637690544128418,393 "learning_rate": 0.00019999638024159827,394 "loss": 4.0287,395 "step": 54396 },397 {398 "epoch": 0.0033833661417322835,399 "grad_norm": 0.6743874549865723,400 "learning_rate": 0.00019999621383846563,401 "loss": 4.4588,402 "step": 55403 },404 {405 "epoch": 0.0034448818897637795,406 "grad_norm": 0.9613650441169739,407 "learning_rate": 0.00019999604369603214,408 "loss": 3.479,409 "step": 56410 },411 {412 "epoch": 0.0035063976377952754,413 "grad_norm": 0.8676753640174866,414 "learning_rate": 0.00019999586981430425,415 "loss": 4.1257,416 "step": 57417 },418 {419 "epoch": 0.003567913385826772,420 "grad_norm": 0.7760907411575317,421 "learning_rate": 0.00019999569219328843,422 "loss": 4.505,423 "step": 58424 },425 {426 "epoch": 0.0036294291338582678,427 "grad_norm": 0.887509822845459,428 "learning_rate": 0.0001999955108329913,429 "loss": 4.8,430 "step": 59431 },432 {433 "epoch": 0.0036909448818897637,434 "grad_norm": 0.7134016752243042,435 "learning_rate": 0.00019999532573341963,436 "loss": 4.7897,437 "step": 60438 },439 {440 "epoch": 0.0037524606299212597,441 "grad_norm": 0.8391125202178955,442 "learning_rate": 0.00019999513689458035,443 "loss": 4.3249,444 "step": 61445 },446 {447 "epoch": 0.003813976377952756,448 "grad_norm": 1.0558634996414185,449 "learning_rate": 0.00019999494431648058,450 "loss": 4.2079,451 "step": 62452 },453 {454 "epoch": 0.003875492125984252,455 "grad_norm": 0.7980897426605225,456 "learning_rate": 0.0001999947479991275,457 "loss": 4.0956,458 "step": 63459 },460 {461 "epoch": 0.003937007874015748,462 "grad_norm": 0.7305802702903748,463 "learning_rate": 0.0001999945479425284,464 "loss": 4.4303,465 "step": 64466 },467 {468 "epoch": 0.003998523622047244,469 "grad_norm": 0.7589819431304932,470 "learning_rate": 0.00019999434414669077,471 "loss": 4.7075,472 "step": 65473 },474 {475 "epoch": 0.00406003937007874,476 "grad_norm": 0.7469328045845032,477 "learning_rate": 0.0001999941366116223,478 "loss": 4.2832,479 "step": 66480 },481 {482 "epoch": 0.004121555118110236,483 "grad_norm": 0.9335817694664001,484 "learning_rate": 0.00019999392533733067,485 "loss": 4.0655,486 "step": 67487 },488 {489 "epoch": 0.004183070866141733,490 "grad_norm": 0.745487630367279,491 "learning_rate": 0.0001999937103238238,492 "loss": 4.3128,493 "step": 68494 },495 {496 "epoch": 0.004244586614173229,497 "grad_norm": 1.0865602493286133,498 "learning_rate": 0.00019999349157110978,499 "loss": 4.8683,500 "step": 69501 },502 {503 "epoch": 0.004306102362204725,504 "grad_norm": 0.6161433458328247,505 "learning_rate": 0.00019999326907919673,506 "loss": 4.3485,507 "step": 70508 },509 {510 "epoch": 0.0043676181102362205,511 "grad_norm": 0.5696821808815002,512 "learning_rate": 0.00019999304284809303,513 "loss": 3.9096,514 "step": 71515 },516 {517 "epoch": 0.0044291338582677165,518 "grad_norm": 0.7559783458709717,519 "learning_rate": 0.00019999281287780708,520 "loss": 3.8548,521 "step": 72522 },523 {524 "epoch": 0.004490649606299212,525 "grad_norm": 0.7913506627082825,526 "learning_rate": 0.0001999925791683475,527 "loss": 4.0762,528 "step": 73529 },530 {531 "epoch": 0.004552165354330708,532 "grad_norm": 0.7013502717018127,533 "learning_rate": 0.000199992341719723,534 "loss": 4.4647,535 "step": 74536 },537 {538 "epoch": 0.004613681102362204,539 "grad_norm": 0.7977866530418396,540 "learning_rate": 0.00019999210053194254,541 "loss": 3.5262,542 "step": 75543 },544 {545 "epoch": 0.004675196850393701,546 "grad_norm": 0.7070642113685608,547 "learning_rate": 0.0001999918556050151,548 "loss": 3.6275,549 "step": 76550 },551 {552 "epoch": 0.004736712598425197,553 "grad_norm": 0.7864336371421814,554 "learning_rate": 0.0001999916069389498,555 "loss": 4.7123,556 "step": 77557 },558 {559 "epoch": 0.004798228346456693,560 "grad_norm": 0.7045543789863586,561 "learning_rate": 0.00019999135453375598,562 "loss": 4.161,563 "step": 78564 },565 {566 "epoch": 0.004859744094488189,567 "grad_norm": 0.9334936738014221,568 "learning_rate": 0.00019999109838944305,569 "loss": 3.781,570 "step": 79571 },572 {573 "epoch": 0.004921259842519685,574 "grad_norm": 0.7052764892578125,575 "learning_rate": 0.00019999083850602063,576 "loss": 3.9291,577 "step": 80578 },579 {580 "epoch": 0.004982775590551181,581 "grad_norm": 0.7392017841339111,582 "learning_rate": 0.0001999905748834984,583 "loss": 4.3447,584 "step": 81585 },586 {587 "epoch": 0.005044291338582677,588 "grad_norm": 0.615789532661438,589 "learning_rate": 0.00019999030752188625,590 "loss": 3.9038,591 "step": 82592 },593 {594 "epoch": 0.005105807086614173,595 "grad_norm": 0.6859673857688904,596 "learning_rate": 0.00019999003642119415,597 "loss": 4.9793,598 "step": 83599 },600 {601 "epoch": 0.00516732283464567,602 "grad_norm": 0.8041189908981323,603 "learning_rate": 0.00019998976158143222,604 "loss": 3.9467,605 "step": 84606 },607 {608 "epoch": 0.005228838582677166,609 "grad_norm": 1.1325663328170776,610 "learning_rate": 0.0001999894830026108,611 "loss": 4.4563,612 "step": 85613 },614 {615 "epoch": 0.0052903543307086616,616 "grad_norm": 1.0158467292785645,617 "learning_rate": 0.00019998920068474028,618 "loss": 4.7595,619 "step": 86620 },621 {622 "epoch": 0.0053518700787401575,623 "grad_norm": 0.9293975234031677,624 "learning_rate": 0.0001999889146278312,625 "loss": 4.3069,626 "step": 87627 },628 {629 "epoch": 0.0054133858267716535,630 "grad_norm": 0.9527814984321594,631 "learning_rate": 0.00019998862483189424,632 "loss": 3.879,633 "step": 88634 },635 {636 "epoch": 0.005474901574803149,637 "grad_norm": 0.8251259326934814,638 "learning_rate": 0.00019998833129694028,639 "loss": 4.5881,640 "step": 89641 },642 {643 "epoch": 0.005536417322834645,644 "grad_norm": 1.355383276939392,645 "learning_rate": 0.00019998803402298027,646 "loss": 4.1265,647 "step": 90648 },649 {650 "epoch": 0.005597933070866141,651 "grad_norm": 0.8191584944725037,652 "learning_rate": 0.00019998773301002537,653 "loss": 3.8795,654 "step": 91655 },656 {657 "epoch": 0.005659448818897638,658 "grad_norm": 0.8508920073509216,659 "learning_rate": 0.00019998742825808676,660 "loss": 3.6558,661 "step": 92662 },663 {664 "epoch": 0.005720964566929134,665 "grad_norm": 0.9597457051277161,666 "learning_rate": 0.0001999871197671759,667 "loss": 4.2718,668 "step": 93669 },670 {671 "epoch": 0.00578248031496063,672 "grad_norm": 1.0464812517166138,673 "learning_rate": 0.0001999868075373043,674 "loss": 4.6338,675 "step": 94676 },677 {678 "epoch": 0.005843996062992126,679 "grad_norm": 1.2142198085784912,680 "learning_rate": 0.00019998649156848365,681 "loss": 3.724,682 "step": 95683 },684 {685 "epoch": 0.005905511811023622,686 "grad_norm": 1.9712352752685547,687 "learning_rate": 0.00019998617186072574,688 "loss": 2.6908,689 "step": 96690 },691 {692 "epoch": 0.005967027559055118,693 "grad_norm": 0.8344887495040894,694 "learning_rate": 0.00019998584841404254,695 "loss": 4.0843,696 "step": 97697 },698 {699 "epoch": 0.006028543307086614,700 "grad_norm": 0.8581575751304626,701 "learning_rate": 0.00019998552122844616,702 "loss": 4.2033,703 "step": 98704 },705 {706 "epoch": 0.00609005905511811,707 "grad_norm": 0.841177225112915,708 "learning_rate": 0.00019998519030394883,709 "loss": 3.5895,710 "step": 99711 },712 {713 "epoch": 0.006151574803149607,714 "grad_norm": 0.8409016132354736,715 "learning_rate": 0.00019998485564056288,716 "loss": 3.6816,717 "step": 100718 },719 {720 "epoch": 0.006213090551181103,721 "grad_norm": 1.1098588705062866,722 "learning_rate": 0.00019998451723830088,723 "loss": 3.6792,724 "step": 101725 },726 {727 "epoch": 0.0062746062992125986,728 "grad_norm": 0.9511058330535889,729 "learning_rate": 0.00019998417509717548,730 "loss": 4.0614,731 "step": 102732 },733 {734 "epoch": 0.0063361220472440945,735 "grad_norm": 0.6864252090454102,736 "learning_rate": 0.00019998382921719942,737 "loss": 3.8433,738 "step": 103739 },740 {741 "epoch": 0.0063976377952755905,742 "grad_norm": 0.8367637395858765,743 "learning_rate": 0.0001999834795983857,744 "loss": 4.4264,745 "step": 104746 },747 {748 "epoch": 0.006459153543307086,749 "grad_norm": 0.8895636796951294,750 "learning_rate": 0.00019998312624074736,751 "loss": 3.8196,752 "step": 105753 },754 {755 "epoch": 0.006520669291338582,756 "grad_norm": 1.2029048204421997,757 "learning_rate": 0.00019998276914429762,758 "loss": 4.642,759 "step": 106760 },761 {762 "epoch": 0.006582185039370078,763 "grad_norm": 0.8605619668960571,764 "learning_rate": 0.00019998240830904984,765 "loss": 3.525,766 "step": 107767 },768 {769 "epoch": 0.006643700787401575,770 "grad_norm": 1.1288233995437622,771 "learning_rate": 0.0001999820437350175,772 "loss": 4.3789,773 "step": 108774 },775 {776 "epoch": 0.006705216535433071,777 "grad_norm": 0.8526942729949951,778 "learning_rate": 0.00019998167542221423,779 "loss": 3.7825,780 "step": 109781 },782 {783 "epoch": 0.006766732283464567,784 "grad_norm": 1.278733491897583,785 "learning_rate": 0.00019998130337065383,786 "loss": 3.7498,787 "step": 110788 },789 {790 "epoch": 0.006828248031496063,791 "grad_norm": 0.7181769609451294,792 "learning_rate": 0.0001999809275803502,793 "loss": 4.2836,794 "step": 111795 },796 {797 "epoch": 0.006889763779527559,798 "grad_norm": 0.9225865006446838,799 "learning_rate": 0.00019998054805131737,800 "loss": 3.9188,801 "step": 112802 },803 {804 "epoch": 0.006951279527559055,805 "grad_norm": 0.9506409168243408,806 "learning_rate": 0.00019998016478356957,807 "loss": 3.8086,808 "step": 113809 },810 {811 "epoch": 0.007012795275590551,812 "grad_norm": 1.0083173513412476,813 "learning_rate": 0.00019997977777712106,814 "loss": 4.0807,815 "step": 114816 },817 {818 "epoch": 0.007074311023622047,819 "grad_norm": 0.8565378189086914,820 "learning_rate": 0.00019997938703198643,821 "loss": 3.7325,822 "step": 115823 },824 {825 "epoch": 0.007135826771653544,826 "grad_norm": 0.7908998131752014,827 "learning_rate": 0.0001999789925481802,828 "loss": 4.0611,829 "step": 116830 },831 {832 "epoch": 0.00719734251968504,833 "grad_norm": 0.7109431028366089,834 "learning_rate": 0.00019997859432571713,835 "loss": 3.9249,836 "step": 117837 },838 {839 "epoch": 0.0072588582677165355,840 "grad_norm": 0.7850291728973389,841 "learning_rate": 0.00019997819236461213,842 "loss": 3.7864,843 "step": 118844 },845 {846 "epoch": 0.0073203740157480315,847 "grad_norm": 1.1817116737365723,848 "learning_rate": 0.00019997778666488024,849 "loss": 4.5303,850 "step": 119851 },852 {853 "epoch": 0.0073818897637795275,854 "grad_norm": 1.043411135673523,855 "learning_rate": 0.00019997737722653663,856 "loss": 3.7504,857 "step": 120858 },859 {860 "epoch": 0.007443405511811023,861 "grad_norm": 0.8229930996894836,862 "learning_rate": 0.0001999769640495966,863 "loss": 4.1244,864 "step": 121865 },866 {867 "epoch": 0.007504921259842519,868 "grad_norm": 0.825101912021637,869 "learning_rate": 0.00019997654713407555,870 "loss": 3.8745,871 "step": 122872 },873 {874 "epoch": 0.007566437007874015,875 "grad_norm": 0.9613546133041382,876 "learning_rate": 0.00019997612647998917,877 "loss": 4.0794,878 "step": 123879 },880 {881 "epoch": 0.007627952755905512,882 "grad_norm": 1.0202386379241943,883 "learning_rate": 0.00019997570208735315,884 "loss": 4.1264,885 "step": 124886 },887 {888 "epoch": 0.007689468503937008,889 "grad_norm": 1.0154190063476562,890 "learning_rate": 0.0001999752739561833,891 "loss": 4.2101,892 "step": 125893 },894 {895 "epoch": 0.007750984251968504,896 "grad_norm": 1.521721363067627,897 "learning_rate": 0.00019997484208649576,898 "loss": 3.69,899 "step": 126900 },901 {902 "epoch": 0.0078125,903 "grad_norm": 1.2424317598342896,904 "learning_rate": 0.00019997440647830653,905 "loss": 3.8559,906 "step": 127907 },908 {909 "epoch": 0.007874015748031496,910 "grad_norm": 0.9340006113052368,911 "learning_rate": 0.000199973967131632,912 "loss": 4.1619,913 "step": 128914 },915 {916 "epoch": 0.007935531496062992,917 "grad_norm": 1.0604355335235596,918 "learning_rate": 0.00019997352404648857,919 "loss": 3.4408,920 "step": 129921 },922 {923 "epoch": 0.007997047244094488,924 "grad_norm": 1.0056222677230835,925 "learning_rate": 0.0001999730772228928,926 "loss": 3.3851,927 "step": 130928 },929 {930 "epoch": 0.008058562992125984,931 "grad_norm": 0.7513048648834229,932 "learning_rate": 0.00019997262666086142,933 "loss": 4.2672,934 "step": 131935 },936 {937 "epoch": 0.00812007874015748,938 "grad_norm": 1.3518322706222534,939 "learning_rate": 0.0001999721723604113,940 "loss": 4.161,941 "step": 132942 },943 {944 "epoch": 0.008181594488188976,945 "grad_norm": 1.1235636472702026,946 "learning_rate": 0.00019997171432155933,947 "loss": 3.9851,948 "step": 133949 },950 {951 "epoch": 0.008243110236220472,952 "grad_norm": 1.1121013164520264,953 "learning_rate": 0.0001999712525443227,954 "loss": 3.3369,955 "step": 134956 },957 {958 "epoch": 0.00830462598425197,959 "grad_norm": 1.1015123128890991,960 "learning_rate": 0.00019997078702871872,961 "loss": 4.1392,962 "step": 135963 },964 {965 "epoch": 0.008366141732283465,966 "grad_norm": 1.2108583450317383,967 "learning_rate": 0.00019997031777476475,968 "loss": 3.5958,969 "step": 136970 },971 {972 "epoch": 0.008427657480314961,973 "grad_norm": 1.0681123733520508,974 "learning_rate": 0.00019996984478247833,975 "loss": 4.0924,976 "step": 137977 },978 {979 "epoch": 0.008489173228346457,980 "grad_norm": 1.279208779335022,981 "learning_rate": 0.00019996936805187718,982 "loss": 3.7196,983 "step": 138984 },985 {986 "epoch": 0.008550688976377953,987 "grad_norm": 1.3705270290374756,988 "learning_rate": 0.00019996888758297909,989 "loss": 4.3487,990 "step": 139991 },992 {993 "epoch": 0.00861220472440945,994 "grad_norm": 0.8328034281730652,995 "learning_rate": 0.00019996840337580207,996 "loss": 3.9334,997 "step": 140998 },999 {1000 "epoch": 0.008673720472440945,1001 "grad_norm": 1.0824898481369019,1002 "learning_rate": 0.0001999679154303642,1003 "loss": 3.9965,1004 "step": 1411005 },1006 {1007 "epoch": 0.008735236220472441,1008 "grad_norm": 1.020643711090088,1009 "learning_rate": 0.00019996742374668372,1010 "loss": 4.4019,1011 "step": 1421012 },1013 {1014 "epoch": 0.008796751968503937,1015 "grad_norm": 0.9669337868690491,1016 "learning_rate": 0.00019996692832477902,1017 "loss": 1.996,1018 "step": 1431019 },1020 {1021 "epoch": 0.008858267716535433,1022 "grad_norm": 0.9339452981948853,1023 "learning_rate": 0.0001999664291646686,1024 "loss": 4.0075,1025 "step": 1441026 },1027 {1028 "epoch": 0.008919783464566929,1029 "grad_norm": 0.9043839573860168,1030 "learning_rate": 0.0001999659262663712,1031 "loss": 3.9553,1032 "step": 1451033 },1034 {1035 "epoch": 0.008981299212598425,1036 "grad_norm": 0.9271431565284729,1037 "learning_rate": 0.00019996541962990559,1038 "loss": 3.8066,1039 "step": 1461040 },1041 {1042 "epoch": 0.00904281496062992,1043 "grad_norm": 1.2157026529312134,1044 "learning_rate": 0.0001999649092552907,1045 "loss": 4.9701,1046 "step": 1471047 },1048 {1049 "epoch": 0.009104330708661417,1050 "grad_norm": 0.9345269203186035,1051 "learning_rate": 0.00019996439514254562,1052 "loss": 3.8809,1053 "step": 1481054 },1055 {1056 "epoch": 0.009165846456692913,1057 "grad_norm": 1.0654510259628296,1058 "learning_rate": 0.00019996387729168954,1059 "loss": 3.9988,1060 "step": 1491061 },1062 {1063 "epoch": 0.009227362204724409,1064 "grad_norm": 1.3758782148361206,1065 "learning_rate": 0.00019996335570274192,1066 "loss": 4.3286,1067 "step": 1501068 },1069 {1070 "epoch": 0.009288877952755906,1071 "grad_norm": 1.23570716381073,1072 "learning_rate": 0.00019996283037572216,1073 "loss": 4.0428,1074 "step": 1511075 },1076 {1077 "epoch": 0.009350393700787402,1078 "grad_norm": 0.9689831733703613,1079 "learning_rate": 0.00019996230131065002,1080 "loss": 4.009,1081 "step": 1521082 },1083 {1084 "epoch": 0.009411909448818898,1085 "grad_norm": 0.9608110189437866,1086 "learning_rate": 0.00019996176850754516,1087 "loss": 3.4991,1088 "step": 1531089 },1090 {1091 "epoch": 0.009473425196850394,1092 "grad_norm": 1.1149580478668213,1093 "learning_rate": 0.00019996123196642756,1094 "loss": 3.2724,1095 "step": 1541096 },1097 {1098 "epoch": 0.00953494094488189,1099 "grad_norm": 1.0518015623092651,1100 "learning_rate": 0.00019996069168731732,1101 "loss": 3.8297,1102 "step": 1551103 },1104 {1105 "epoch": 0.009596456692913386,1106 "grad_norm": 1.4498533010482788,1107 "learning_rate": 0.00019996014767023458,1108 "loss": 4.0394,1109 "step": 1561110 },1111 {1112 "epoch": 0.009657972440944882,1113 "grad_norm": 0.9268041253089905,1114 "learning_rate": 0.0001999595999151997,1115 "loss": 3.5148,1116 "step": 1571117 },1118 {1119 "epoch": 0.009719488188976378,1120 "grad_norm": 0.9180638194084167,1121 "learning_rate": 0.0001999590484222332,1122 "loss": 3.4739,1123 "step": 1581124 },1125 {1126 "epoch": 0.009781003937007874,1127 "grad_norm": 1.0417743921279907,1128 "learning_rate": 0.00019995849319135567,1129 "loss": 4.1415,1130 "step": 1591131 },1132 {1133 "epoch": 0.00984251968503937,1134 "grad_norm": 0.8904036283493042,1135 "learning_rate": 0.00019995793422258786,1136 "loss": 4.0283,1137 "step": 1601138 },1139 {1140 "epoch": 0.009904035433070866,1141 "grad_norm": 1.0201314687728882,1142 "learning_rate": 0.00019995737151595072,1143 "loss": 4.0576,1144 "step": 1611145 },1146 {1147 "epoch": 0.009965551181102362,1148 "grad_norm": 1.109643578529358,1149 "learning_rate": 0.00019995680507146522,1150 "loss": 3.6832,1151 "step": 1621152 },1153 {1154 "epoch": 0.010027066929133858,1155 "grad_norm": 0.8940156102180481,1156 "learning_rate": 0.0001999562348891526,1157 "loss": 3.6519,1158 "step": 1631159 },1160 {1161 "epoch": 0.010088582677165354,1162 "grad_norm": 1.2682384252548218,1163 "learning_rate": 0.0001999556609690342,1164 "loss": 4.2614,1165 "step": 1641166 },1167 {1168 "epoch": 0.01015009842519685,1169 "grad_norm": 1.113534688949585,1170 "learning_rate": 0.0001999550833111314,1171 "loss": 3.2479,1172 "step": 1651173 },1174 {1175 "epoch": 0.010211614173228346,1176 "grad_norm": 1.5006425380706787,1177 "learning_rate": 0.00019995450191546588,1178 "loss": 3.6487,1179 "step": 1661180 },1181 {1182 "epoch": 0.010273129921259843,1183 "grad_norm": 1.1574208736419678,1184 "learning_rate": 0.00019995391678205934,1185 "loss": 3.6985,1186 "step": 1671187 },1188 {1189 "epoch": 0.01033464566929134,1190 "grad_norm": 0.9824999570846558,1191 "learning_rate": 0.0001999533279109337,1192 "loss": 3.7214,1193 "step": 1681194 },1195 {1196 "epoch": 0.010396161417322835,1197 "grad_norm": 0.9681886434555054,1198 "learning_rate": 0.0001999527353021109,1199 "loss": 3.8614,1200 "step": 169