Azrail/smallm_70_rope
043
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 0.4708549211906576,6 "eval_steps": 500,7 "global_step": 70000,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.00033632494370761253,14 "grad_norm": 1.274348497390747,15 "learning_rate": 0.0001,16 "loss": 9.9462,17 "num_input_tokens_seen": 13107200,18 "step": 5019 },20 {21 "epoch": 0.0006726498874152251,22 "grad_norm": 0.3978561758995056,23 "learning_rate": 0.0002,24 "loss": 8.361,25 "num_input_tokens_seen": 26214400,26 "step": 10027 },28 {29 "epoch": 0.0010089748311228376,30 "grad_norm": 0.41306379437446594,31 "learning_rate": 0.0003,32 "loss": 7.6593,33 "num_input_tokens_seen": 39321600,34 "step": 15035 },36 {37 "epoch": 0.0013452997748304501,38 "grad_norm": 0.3370315134525299,39 "learning_rate": 0.0004,40 "loss": 7.1018,41 "num_input_tokens_seen": 52428800,42 "step": 20043 },44 {45 "epoch": 0.0016816247185380627,46 "grad_norm": 0.6396230459213257,47 "learning_rate": 0.0005,48 "loss": 6.6331,49 "num_input_tokens_seen": 65536000,50 "step": 25051 },52 {53 "epoch": 0.002017949662245675,54 "grad_norm": 0.5036832690238953,55 "learning_rate": 0.0006,56 "loss": 6.2646,57 "num_input_tokens_seen": 78643200,58 "step": 30059 },60 {61 "epoch": 0.002354274605953288,62 "grad_norm": 0.4829367399215698,63 "learning_rate": 0.0007,64 "loss": 5.9656,65 "num_input_tokens_seen": 91750400,66 "step": 35067 },68 {69 "epoch": 0.0026905995496609002,70 "grad_norm": 0.6319091320037842,71 "learning_rate": 0.0008,72 "loss": 5.7098,73 "num_input_tokens_seen": 104857600,74 "step": 40075 },76 {77 "epoch": 0.003026924493368513,78 "grad_norm": 0.8607974648475647,79 "learning_rate": 0.0009000000000000001,80 "loss": 5.5012,81 "num_input_tokens_seen": 117964800,82 "step": 45083 },84 {85 "epoch": 0.0033632494370761253,86 "grad_norm": 0.4290812611579895,87 "learning_rate": 0.001,88 "loss": 5.3179,89 "num_input_tokens_seen": 131072000,90 "step": 50091 },92 {93 "epoch": 0.0033632494370761253,94 "eval_loss": 5.179312705993652,95 "eval_runtime": 100.6159,96 "eval_samples_per_second": 49.694,97 "eval_steps_per_second": 12.423,98 "num_input_tokens_seen": 131072000,99 "step": 500100 },101 {102 "epoch": 0.003699574380783738,103 "grad_norm": 0.407632052898407,104 "learning_rate": 0.001,105 "loss": 5.1431,106 "num_input_tokens_seen": 144179200,107 "step": 550108 },109 {110 "epoch": 0.00403589932449135,111 "grad_norm": 0.31742918491363525,112 "learning_rate": 0.001,113 "loss": 4.9611,114 "num_input_tokens_seen": 157286400,115 "step": 600116 },117 {118 "epoch": 0.004372224268198963,119 "grad_norm": 0.3897629678249359,120 "learning_rate": 0.001,121 "loss": 4.8043,122 "num_input_tokens_seen": 170393600,123 "step": 650124 },125 {126 "epoch": 0.004708549211906576,127 "grad_norm": 0.41930389404296875,128 "learning_rate": 0.001,129 "loss": 4.6617,130 "num_input_tokens_seen": 183500800,131 "step": 700132 },133 {134 "epoch": 0.005044874155614189,135 "grad_norm": 0.4283705949783325,136 "learning_rate": 0.001,137 "loss": 4.5535,138 "num_input_tokens_seen": 196608000,139 "step": 750140 },141 {142 "epoch": 0.0053811990993218005,143 "grad_norm": 0.3931107521057129,144 "learning_rate": 0.001,145 "loss": 4.4556,146 "num_input_tokens_seen": 209715200,147 "step": 800148 },149 {150 "epoch": 0.005717524043029413,151 "grad_norm": 0.3258611857891083,152 "learning_rate": 0.001,153 "loss": 4.3808,154 "num_input_tokens_seen": 222822400,155 "step": 850156 },157 {158 "epoch": 0.006053848986737026,159 "grad_norm": 0.3562588095664978,160 "learning_rate": 0.001,161 "loss": 4.3136,162 "num_input_tokens_seen": 235929600,163 "step": 900164 },165 {166 "epoch": 0.006390173930444639,167 "grad_norm": 0.3626460134983063,168 "learning_rate": 0.001,169 "loss": 4.2557,170 "num_input_tokens_seen": 249036800,171 "step": 950172 },173 {174 "epoch": 0.006726498874152251,175 "grad_norm": 0.4001849293708801,176 "learning_rate": 0.001,177 "loss": 4.208,178 "num_input_tokens_seen": 262144000,179 "step": 1000180 },181 {182 "epoch": 0.006726498874152251,183 "eval_loss": 4.108783721923828,184 "eval_runtime": 101.5418,185 "eval_samples_per_second": 49.241,186 "eval_steps_per_second": 12.31,187 "num_input_tokens_seen": 262144000,188 "step": 1000189 },190 {191 "epoch": 0.007062823817859863,192 "grad_norm": 0.36548155546188354,193 "learning_rate": 0.001,194 "loss": 4.1537,195 "num_input_tokens_seen": 275251200,196 "step": 1050197 },198 {199 "epoch": 0.007399148761567476,200 "grad_norm": 0.33794787526130676,201 "learning_rate": 0.001,202 "loss": 4.1062,203 "num_input_tokens_seen": 288358400,204 "step": 1100205 },206 {207 "epoch": 0.007735473705275089,208 "grad_norm": 0.3374481499195099,209 "learning_rate": 0.001,210 "loss": 4.0703,211 "num_input_tokens_seen": 301465600,212 "step": 1150213 },214 {215 "epoch": 0.0080717986489827,216 "grad_norm": 0.3061329126358032,217 "learning_rate": 0.001,218 "loss": 4.0253,219 "num_input_tokens_seen": 314572800,220 "step": 1200221 },222 {223 "epoch": 0.008408123592690313,224 "grad_norm": 0.3140158951282501,225 "learning_rate": 0.001,226 "loss": 4.0128,227 "num_input_tokens_seen": 327680000,228 "step": 1250229 },230 {231 "epoch": 0.008744448536397926,232 "grad_norm": 0.3002954125404358,233 "learning_rate": 0.001,234 "loss": 3.9762,235 "num_input_tokens_seen": 340787200,236 "step": 1300237 },238 {239 "epoch": 0.009080773480105539,240 "grad_norm": 0.2994467318058014,241 "learning_rate": 0.001,242 "loss": 3.9485,243 "num_input_tokens_seen": 353894400,244 "step": 1350245 },246 {247 "epoch": 0.009417098423813152,248 "grad_norm": 0.25649985671043396,249 "learning_rate": 0.001,250 "loss": 3.9298,251 "num_input_tokens_seen": 367001600,252 "step": 1400253 },254 {255 "epoch": 0.009753423367520764,256 "grad_norm": 0.2627107501029968,257 "learning_rate": 0.001,258 "loss": 3.904,259 "num_input_tokens_seen": 380108800,260 "step": 1450261 },262 {263 "epoch": 0.010089748311228377,264 "grad_norm": 0.29527419805526733,265 "learning_rate": 0.001,266 "loss": 3.8864,267 "num_input_tokens_seen": 393216000,268 "step": 1500269 },270 {271 "epoch": 0.010089748311228377,272 "eval_loss": 3.8075473308563232,273 "eval_runtime": 99.5484,274 "eval_samples_per_second": 50.227,275 "eval_steps_per_second": 12.557,276 "num_input_tokens_seen": 393216000,277 "step": 1500278 },279 {280 "epoch": 0.010426073254935988,281 "grad_norm": 0.26328331232070923,282 "learning_rate": 0.001,283 "loss": 3.8701,284 "num_input_tokens_seen": 406323200,285 "step": 1550286 },287 {288 "epoch": 0.010762398198643601,289 "grad_norm": 0.2574864625930786,290 "learning_rate": 0.001,291 "loss": 3.8575,292 "num_input_tokens_seen": 419430400,293 "step": 1600294 },295 {296 "epoch": 0.011098723142351214,297 "grad_norm": 0.2445235550403595,298 "learning_rate": 0.001,299 "loss": 3.8277,300 "num_input_tokens_seen": 432537600,301 "step": 1650302 },303 {304 "epoch": 0.011435048086058826,305 "grad_norm": 0.2726516127586365,306 "learning_rate": 0.001,307 "loss": 3.8171,308 "num_input_tokens_seen": 445644800,309 "step": 1700310 },311 {312 "epoch": 0.01177137302976644,313 "grad_norm": 0.2658848166465759,314 "learning_rate": 0.001,315 "loss": 3.8026,316 "num_input_tokens_seen": 458752000,317 "step": 1750318 },319 {320 "epoch": 0.012107697973474052,321 "grad_norm": 0.30301713943481445,322 "learning_rate": 0.001,323 "loss": 3.7838,324 "num_input_tokens_seen": 471859200,325 "step": 1800326 },327 {328 "epoch": 0.012444022917181665,329 "grad_norm": 0.24884650111198425,330 "learning_rate": 0.001,331 "loss": 3.7685,332 "num_input_tokens_seen": 484966400,333 "step": 1850334 },335 {336 "epoch": 0.012780347860889277,337 "grad_norm": 0.2728760540485382,338 "learning_rate": 0.001,339 "loss": 3.759,340 "num_input_tokens_seen": 498073600,341 "step": 1900342 },343 {344 "epoch": 0.013116672804596888,345 "grad_norm": 0.24091999232769012,346 "learning_rate": 0.001,347 "loss": 3.7496,348 "num_input_tokens_seen": 511180800,349 "step": 1950350 },351 {352 "epoch": 0.013452997748304501,353 "grad_norm": 0.2559104561805725,354 "learning_rate": 0.001,355 "loss": 3.7289,356 "num_input_tokens_seen": 524288000,357 "step": 2000358 },359 {360 "epoch": 0.013452997748304501,361 "eval_loss": 3.654578447341919,362 "eval_runtime": 99.6741,363 "eval_samples_per_second": 50.164,364 "eval_steps_per_second": 12.541,365 "num_input_tokens_seen": 524288000,366 "step": 2000367 },368 {369 "epoch": 0.013789322692012114,370 "grad_norm": 0.24713733792304993,371 "learning_rate": 0.001,372 "loss": 3.7259,373 "num_input_tokens_seen": 537395200,374 "step": 2050375 },376 {377 "epoch": 0.014125647635719727,378 "grad_norm": 0.2594991624355316,379 "learning_rate": 0.001,380 "loss": 3.7121,381 "num_input_tokens_seen": 550502400,382 "step": 2100383 },384 {385 "epoch": 0.01446197257942734,386 "grad_norm": 0.2594399154186249,387 "learning_rate": 0.001,388 "loss": 3.6975,389 "num_input_tokens_seen": 563609600,390 "step": 2150391 },392 {393 "epoch": 0.014798297523134952,394 "grad_norm": 0.21877750754356384,395 "learning_rate": 0.001,396 "loss": 3.6868,397 "num_input_tokens_seen": 576716800,398 "step": 2200399 },400 {401 "epoch": 0.015134622466842565,402 "grad_norm": 0.23657166957855225,403 "learning_rate": 0.001,404 "loss": 3.6853,405 "num_input_tokens_seen": 589824000,406 "step": 2250407 },408 {409 "epoch": 0.015470947410550178,410 "grad_norm": 0.24121630191802979,411 "learning_rate": 0.001,412 "loss": 3.6718,413 "num_input_tokens_seen": 602931200,414 "step": 2300415 },416 {417 "epoch": 0.01580727235425779,418 "grad_norm": 0.25701209902763367,419 "learning_rate": 0.001,420 "loss": 3.6658,421 "num_input_tokens_seen": 616038400,422 "step": 2350423 },424 {425 "epoch": 0.0161435972979654,426 "grad_norm": 0.22402645647525787,427 "learning_rate": 0.001,428 "loss": 3.6704,429 "num_input_tokens_seen": 629145600,430 "step": 2400431 },432 {433 "epoch": 0.016479922241673016,434 "grad_norm": 0.2358316332101822,435 "learning_rate": 0.001,436 "loss": 3.6464,437 "num_input_tokens_seen": 642252800,438 "step": 2450439 },440 {441 "epoch": 0.016816247185380627,442 "grad_norm": 0.23835696280002594,443 "learning_rate": 0.001,444 "loss": 3.6424,445 "num_input_tokens_seen": 655360000,446 "step": 2500447 },448 {449 "epoch": 0.016816247185380627,450 "eval_loss": 3.560713768005371,451 "eval_runtime": 100.1074,452 "eval_samples_per_second": 49.946,453 "eval_steps_per_second": 12.487,454 "num_input_tokens_seen": 655360000,455 "step": 2500456 },457 {458 "epoch": 0.01715257212908824,459 "grad_norm": 0.20949696004390717,460 "learning_rate": 0.001,461 "loss": 3.6346,462 "num_input_tokens_seen": 668467200,463 "step": 2550464 },465 {466 "epoch": 0.017488897072795852,467 "grad_norm": 0.21765078604221344,468 "learning_rate": 0.001,469 "loss": 3.6123,470 "num_input_tokens_seen": 681574400,471 "step": 2600472 },473 {474 "epoch": 0.017825222016503463,475 "grad_norm": 0.21167752146720886,476 "learning_rate": 0.001,477 "loss": 3.6173,478 "num_input_tokens_seen": 694681600,479 "step": 2650480 },481 {482 "epoch": 0.018161546960211078,483 "grad_norm": 0.22060410678386688,484 "learning_rate": 0.001,485 "loss": 3.62,486 "num_input_tokens_seen": 707788800,487 "step": 2700488 },489 {490 "epoch": 0.01849787190391869,491 "grad_norm": 0.21778976917266846,492 "learning_rate": 0.001,493 "loss": 3.6009,494 "num_input_tokens_seen": 720896000,495 "step": 2750496 },497 {498 "epoch": 0.018834196847626303,499 "grad_norm": 0.21130047738552094,500 "learning_rate": 0.001,501 "loss": 3.5882,502 "num_input_tokens_seen": 734003200,503 "step": 2800504 },505 {506 "epoch": 0.019170521791333914,507 "grad_norm": 0.20137132704257965,508 "learning_rate": 0.001,509 "loss": 3.5923,510 "num_input_tokens_seen": 747110400,511 "step": 2850512 },513 {514 "epoch": 0.01950684673504153,515 "grad_norm": 0.24937641620635986,516 "learning_rate": 0.001,517 "loss": 3.5872,518 "num_input_tokens_seen": 760217600,519 "step": 2900520 },521 {522 "epoch": 0.01984317167874914,523 "grad_norm": 0.20992155373096466,524 "learning_rate": 0.001,525 "loss": 3.5786,526 "num_input_tokens_seen": 773324800,527 "step": 2950528 },529 {530 "epoch": 0.020179496622456754,531 "grad_norm": 0.24723300337791443,532 "learning_rate": 0.001,533 "loss": 3.5846,534 "num_input_tokens_seen": 786432000,535 "step": 3000536 },537 {538 "epoch": 0.020179496622456754,539 "eval_loss": 3.502939462661743,540 "eval_runtime": 99.6836,541 "eval_samples_per_second": 50.159,542 "eval_steps_per_second": 12.54,543 "num_input_tokens_seen": 786432000,544 "step": 3000545 },546 {547 "epoch": 0.020515821566164365,548 "grad_norm": 0.22368234395980835,549 "learning_rate": 0.001,550 "loss": 3.5744,551 "num_input_tokens_seen": 799539200,552 "step": 3050553 },554 {555 "epoch": 0.020852146509871976,556 "grad_norm": 0.20934642851352692,557 "learning_rate": 0.001,558 "loss": 3.5666,559 "num_input_tokens_seen": 812646400,560 "step": 3100561 },562 {563 "epoch": 0.02118847145357959,564 "grad_norm": 0.1929185688495636,565 "learning_rate": 0.001,566 "loss": 3.5624,567 "num_input_tokens_seen": 825753600,568 "step": 3150569 },570 {571 "epoch": 0.021524796397287202,572 "grad_norm": 0.20416973531246185,573 "learning_rate": 0.001,574 "loss": 3.5491,575 "num_input_tokens_seen": 838860800,576 "step": 3200577 },578 {579 "epoch": 0.021861121340994816,580 "grad_norm": 0.20873814821243286,581 "learning_rate": 0.001,582 "loss": 3.5509,583 "num_input_tokens_seen": 851968000,584 "step": 3250585 },586 {587 "epoch": 0.022197446284702427,588 "grad_norm": 0.22235794365406036,589 "learning_rate": 0.001,590 "loss": 3.5453,591 "num_input_tokens_seen": 865075200,592 "step": 3300593 },594 {595 "epoch": 0.022533771228410042,596 "grad_norm": 0.2174178808927536,597 "learning_rate": 0.001,598 "loss": 3.5378,599 "num_input_tokens_seen": 878182400,600 "step": 3350601 },602 {603 "epoch": 0.022870096172117653,604 "grad_norm": 0.2016495019197464,605 "learning_rate": 0.001,606 "loss": 3.5352,607 "num_input_tokens_seen": 891289600,608 "step": 3400609 },610 {611 "epoch": 0.023206421115825267,612 "grad_norm": 0.1864960640668869,613 "learning_rate": 0.001,614 "loss": 3.524,615 "num_input_tokens_seen": 904396800,616 "step": 3450617 },618 {619 "epoch": 0.02354274605953288,620 "grad_norm": 0.19614097476005554,621 "learning_rate": 0.001,622 "loss": 3.528,623 "num_input_tokens_seen": 917504000,624 "step": 3500625 },626 {627 "epoch": 0.02354274605953288,628 "eval_loss": 3.4472692012786865,629 "eval_runtime": 101.0365,630 "eval_samples_per_second": 49.487,631 "eval_steps_per_second": 12.372,632 "num_input_tokens_seen": 917504000,633 "step": 3500634 },635 {636 "epoch": 0.02387907100324049,637 "grad_norm": 0.18501116335391998,638 "learning_rate": 0.001,639 "loss": 3.53,640 "num_input_tokens_seen": 930611200,641 "step": 3550642 },643 {644 "epoch": 0.024215395946948104,645 "grad_norm": 0.1863412857055664,646 "learning_rate": 0.001,647 "loss": 3.5145,648 "num_input_tokens_seen": 943718400,649 "step": 3600650 },651 {652 "epoch": 0.024551720890655715,653 "grad_norm": 0.1979917287826538,654 "learning_rate": 0.001,655 "loss": 3.5143,656 "num_input_tokens_seen": 956825600,657 "step": 3650658 },659 {660 "epoch": 0.02488804583436333,661 "grad_norm": 0.1991748809814453,662 "learning_rate": 0.001,663 "loss": 3.5065,664 "num_input_tokens_seen": 969932800,665 "step": 3700666 },667 {668 "epoch": 0.02522437077807094,669 "grad_norm": 0.19475233554840088,670 "learning_rate": 0.001,671 "loss": 3.5011,672 "num_input_tokens_seen": 983040000,673 "step": 3750674 },675 {676 "epoch": 0.025560695721778555,677 "grad_norm": 0.195469468832016,678 "learning_rate": 0.001,679 "loss": 3.506,680 "num_input_tokens_seen": 996147200,681 "step": 3800682 },683 {684 "epoch": 0.025897020665486166,685 "grad_norm": 0.19666293263435364,686 "learning_rate": 0.001,687 "loss": 3.4936,688 "num_input_tokens_seen": 1009254400,689 "step": 3850690 },691 {692 "epoch": 0.026233345609193777,693 "grad_norm": 0.20198987424373627,694 "learning_rate": 0.001,695 "loss": 3.4873,696 "num_input_tokens_seen": 1022361600,697 "step": 3900698 },699 {700 "epoch": 0.02656967055290139,701 "grad_norm": 0.18537157773971558,702 "learning_rate": 0.001,703 "loss": 3.4939,704 "num_input_tokens_seen": 1035468800,705 "step": 3950706 },707 {708 "epoch": 0.026905995496609002,709 "grad_norm": 0.18743236362934113,710 "learning_rate": 0.001,711 "loss": 3.4784,712 "num_input_tokens_seen": 1048576000,713 "step": 4000714 },715 {716 "epoch": 0.026905995496609002,717 "eval_loss": 3.4036922454833984,718 "eval_runtime": 99.4277,719 "eval_samples_per_second": 50.288,720 "eval_steps_per_second": 12.572,721 "num_input_tokens_seen": 1048576000,722 "step": 4000723 },724 {725 "epoch": 0.027242320440316617,726 "grad_norm": 0.18962019681930542,727 "learning_rate": 0.001,728 "loss": 3.4848,729 "num_input_tokens_seen": 1061683200,730 "step": 4050731 },732 {733 "epoch": 0.027578645384024228,734 "grad_norm": 0.19415706396102905,735 "learning_rate": 0.001,736 "loss": 3.4934,737 "num_input_tokens_seen": 1074790400,738 "step": 4100739 },740 {741 "epoch": 0.027914970327731842,742 "grad_norm": 0.17727437615394592,743 "learning_rate": 0.001,744 "loss": 3.4835,745 "num_input_tokens_seen": 1087897600,746 "step": 4150747 },748 {749 "epoch": 0.028251295271439453,750 "grad_norm": 0.18400466442108154,751 "learning_rate": 0.001,752 "loss": 3.4843,753 "num_input_tokens_seen": 1101004800,754 "step": 4200755 },756 {757 "epoch": 0.028587620215147068,758 "grad_norm": 0.19114799797534943,759 "learning_rate": 0.001,760 "loss": 3.4713,761 "num_input_tokens_seen": 1114112000,762 "step": 4250763 },764 {765 "epoch": 0.02892394515885468,766 "grad_norm": 0.18681153655052185,767 "learning_rate": 0.001,768 "loss": 3.4601,769 "num_input_tokens_seen": 1127219200,770 "step": 4300771 },772 {773 "epoch": 0.02926027010256229,774 "grad_norm": 0.20739078521728516,775 "learning_rate": 0.001,776 "loss": 3.4628,777 "num_input_tokens_seen": 1140326400,778 "step": 4350779 },780 {781 "epoch": 0.029596595046269904,782 "grad_norm": 0.18018484115600586,783 "learning_rate": 0.001,784 "loss": 3.4521,785 "num_input_tokens_seen": 1153433600,786 "step": 4400787 },788 {789 "epoch": 0.029932919989977515,790 "grad_norm": 0.18144090473651886,791 "learning_rate": 0.001,792 "loss": 3.4536,793 "num_input_tokens_seen": 1166540800,794 "step": 4450795 },796 {797 "epoch": 0.03026924493368513,798 "grad_norm": 0.17444822192192078,799 "learning_rate": 0.001,800 "loss": 3.4509,801 "num_input_tokens_seen": 1179648000,802 "step": 4500803 },804 {805 "epoch": 0.03026924493368513,806 "eval_loss": 3.3682761192321777,807 "eval_runtime": 100.0421,808 "eval_samples_per_second": 49.979,809 "eval_steps_per_second": 12.495,810 "num_input_tokens_seen": 1179648000,811 "step": 4500812 },813 {814 "epoch": 0.03060556987739274,815 "grad_norm": 0.21026909351348877,816 "learning_rate": 0.001,817 "loss": 3.4493,818 "num_input_tokens_seen": 1192755200,819 "step": 4550820 },821 {822 "epoch": 0.030941894821100355,823 "grad_norm": 0.1758560836315155,824 "learning_rate": 0.001,825 "loss": 3.4416,826 "num_input_tokens_seen": 1205862400,827 "step": 4600828 },829 {830 "epoch": 0.03127821976480797,831 "grad_norm": 0.1979188174009323,832 "learning_rate": 0.001,833 "loss": 3.4353,834 "num_input_tokens_seen": 1218969600,835 "step": 4650836 },837 {838 "epoch": 0.03161454470851558,839 "grad_norm": 0.17621161043643951,840 "learning_rate": 0.001,841 "loss": 3.435,842 "num_input_tokens_seen": 1232076800,843 "step": 4700844 },845 {846 "epoch": 0.03195086965222319,847 "grad_norm": 0.18691854178905487,848 "learning_rate": 0.001,849 "loss": 3.4367,850 "num_input_tokens_seen": 1245184000,851 "step": 4750852 },853 {854 "epoch": 0.0322871945959308,855 "grad_norm": 0.17202869057655334,856 "learning_rate": 0.001,857 "loss": 3.4426,858 "num_input_tokens_seen": 1258291200,859 "step": 4800860 },861 {862 "epoch": 0.032623519539638414,863 "grad_norm": 0.17097769677639008,864 "learning_rate": 0.001,865 "loss": 3.439,866 "num_input_tokens_seen": 1271398400,867 "step": 4850868 },869 {870 "epoch": 0.03295984448334603,871 "grad_norm": 0.1845879703760147,872 "learning_rate": 0.001,873 "loss": 3.4263,874 "num_input_tokens_seen": 1284505600,875 "step": 4900876 },877 {878 "epoch": 0.03329616942705364,879 "grad_norm": 0.18462544679641724,880 "learning_rate": 0.001,881 "loss": 3.4273,882 "num_input_tokens_seen": 1297612800,883 "step": 4950884 },885 {886 "epoch": 0.033632494370761254,887 "grad_norm": 0.18930740654468536,888 "learning_rate": 0.001,889 "loss": 3.4252,890 "num_input_tokens_seen": 1310720000,891 "step": 5000892 },893 {894 "epoch": 0.033632494370761254,895 "eval_loss": 3.3413195610046387,896 "eval_runtime": 101.3099,897 "eval_samples_per_second": 49.354,898 "eval_steps_per_second": 12.338,899 "num_input_tokens_seen": 1310720000,900 "step": 5000901 },902 {903 "epoch": 0.033968819314468865,904 "grad_norm": 0.1741451472043991,905 "learning_rate": 0.001,906 "loss": 3.4216,907 "num_input_tokens_seen": 1323827200,908 "step": 5050909 },910 {911 "epoch": 0.03430514425817648,912 "grad_norm": 0.18628782033920288,913 "learning_rate": 0.001,914 "loss": 3.418,915 "num_input_tokens_seen": 1336934400,916 "step": 5100917 },918 {919 "epoch": 0.034641469201884094,920 "grad_norm": 0.17286072671413422,921 "learning_rate": 0.001,922 "loss": 3.4126,923 "num_input_tokens_seen": 1350041600,924 "step": 5150925 },926 {927 "epoch": 0.034977794145591705,928 "grad_norm": 0.18905507028102875,929 "learning_rate": 0.001,930 "loss": 3.4169,931 "num_input_tokens_seen": 1363148800,932 "step": 5200933 },934 {935 "epoch": 0.035314119089299316,936 "grad_norm": 0.17675209045410156,937 "learning_rate": 0.001,938 "loss": 3.4229,939 "num_input_tokens_seen": 1376256000,940 "step": 5250941 },942 {943 "epoch": 0.03565044403300693,944 "grad_norm": 0.19063080847263336,945 "learning_rate": 0.001,946 "loss": 3.4099,947 "num_input_tokens_seen": 1389363200,948 "step": 5300949 },950 {951 "epoch": 0.035986768976714545,952 "grad_norm": 0.1875571757555008,953 "learning_rate": 0.001,954 "loss": 3.4185,955 "num_input_tokens_seen": 1402470400,956 "step": 5350957 },958 {959 "epoch": 0.036323093920422156,960 "grad_norm": 1.0721184015274048,961 "learning_rate": 0.001,962 "loss": 3.4298,963 "num_input_tokens_seen": 1415577600,964 "step": 5400965 },966 {967 "epoch": 0.03665941886412977,968 "grad_norm": 0.1907675564289093,969 "learning_rate": 0.001,970 "loss": 3.4268,971 "num_input_tokens_seen": 1428684800,972 "step": 5450973 },974 {975 "epoch": 0.03699574380783738,976 "grad_norm": 0.18285758793354034,977 "learning_rate": 0.001,978 "loss": 3.4036,979 "num_input_tokens_seen": 1441792000,980 "step": 5500981 },982 {983 "epoch": 0.03699574380783738,984 "eval_loss": 3.31872820854187,985 "eval_runtime": 101.3489,986 "eval_samples_per_second": 49.335,987 "eval_steps_per_second": 12.334,988 "num_input_tokens_seen": 1441792000,989 "step": 5500990 },991 {992 "epoch": 0.037332068751544996,993 "grad_norm": 0.18330691754817963,994 "learning_rate": 0.001,995 "loss": 3.3926,996 "num_input_tokens_seen": 1454899200,997 "step": 5550998 },999 {1000 "epoch": 0.03766839369525261,1001 "grad_norm": 0.16875725984573364,1002 "learning_rate": 0.001,1003 "loss": 3.3975,1004 "num_input_tokens_seen": 1468006400,1005 "step": 56001006 },1007 {1008 "epoch": 0.03800471863896022,1009 "grad_norm": 0.18510381877422333,1010 "learning_rate": 0.001,1011 "loss": 3.3964,1012 "num_input_tokens_seen": 1481113600,1013 "step": 56501014 },1015 {1016 "epoch": 0.03834104358266783,1017 "grad_norm": 0.16602838039398193,1018 "learning_rate": 0.001,1019 "loss": 3.3819,1020 "num_input_tokens_seen": 1494220800,1021 "step": 57001022 },1023 {1024 "epoch": 0.03867736852637544,1025 "grad_norm": 0.16771391034126282,1026 "learning_rate": 0.001,1027 "loss": 3.3968,1028 "num_input_tokens_seen": 1507328000,1029 "step": 57501030 },1031 {1032 "epoch": 0.03901369347008306,1033 "grad_norm": 0.16801221668720245,1034 "learning_rate": 0.001,1035 "loss": 3.3921,1036 "num_input_tokens_seen": 1520435200,1037 "step": 58001038 },1039 {1040 "epoch": 0.03935001841379067,1041 "grad_norm": 0.16846245527267456,1042 "learning_rate": 0.001,1043 "loss": 3.3841,1044 "num_input_tokens_seen": 1533542400,1045 "step": 58501046 },1047 {1048 "epoch": 0.03968634335749828,1049 "grad_norm": 0.17359821498394012,1050 "learning_rate": 0.001,1051 "loss": 3.372,1052 "num_input_tokens_seen": 1546649600,1053 "step": 59001054 },1055 {1056 "epoch": 0.04002266830120589,1057 "grad_norm": 0.1578226536512375,1058 "learning_rate": 0.001,1059 "loss": 3.3826,1060 "num_input_tokens_seen": 1559756800,1061 "step": 59501062 },1063 {1064 "epoch": 0.04035899324491351,1065 "grad_norm": 0.17228208482265472,1066 "learning_rate": 0.001,1067 "loss": 3.3953,1068 "num_input_tokens_seen": 1572864000,1069 "step": 60001070 },1071 {1072 "epoch": 0.04035899324491351,1073 "eval_loss": 3.293433666229248,1074 "eval_runtime": 100.6763,1075 "eval_samples_per_second": 49.664,1076 "eval_steps_per_second": 12.416,1077 "num_input_tokens_seen": 1572864000,1078 "step": 60001079 },1080 {1081 "epoch": 0.04069531818862112,1082 "grad_norm": 0.18091177940368652,1083 "learning_rate": 0.001,1084 "loss": 3.3783,1085 "num_input_tokens_seen": 1585971200,1086 "step": 60501087 },1088 {1089 "epoch": 0.04103164313232873,1090 "grad_norm": 0.17565099895000458,1091 "learning_rate": 0.001,1092 "loss": 3.3773,1093 "num_input_tokens_seen": 1599078400,1094 "step": 61001095 },1096 {1097 "epoch": 0.04136796807603634,1098 "grad_norm": 0.1635759323835373,1099 "learning_rate": 0.001,1100 "loss": 3.3823,1101 "num_input_tokens_seen": 1612185600,1102 "step": 61501103 },1104 {1105 "epoch": 0.04170429301974395,1106 "grad_norm": 0.19144974648952484,1107 "learning_rate": 0.001,1108 "loss": 3.3665,1109 "num_input_tokens_seen": 1625292800,1110 "step": 62001111 },1112 {1113 "epoch": 0.04204061796345157,1114 "grad_norm": 0.1741226762533188,1115 "learning_rate": 0.001,1116 "loss": 3.3637,1117 "num_input_tokens_seen": 1638400000,1118 "step": 62501119 },1120 {1121 "epoch": 0.04237694290715918,1122 "grad_norm": 0.17072845995426178,1123 "learning_rate": 0.001,1124 "loss": 3.3627,1125 "num_input_tokens_seen": 1651507200,1126 "step": 63001127 },1128 {1129 "epoch": 0.04271326785086679,1130 "grad_norm": 0.16942182183265686,1131 "learning_rate": 0.001,1132 "loss": 3.3729,1133 "num_input_tokens_seen": 1664614400,1134 "step": 63501135 },1136 {1137 "epoch": 0.043049592794574404,1138 "grad_norm": 0.16412265598773956,1139 "learning_rate": 0.001,1140 "loss": 3.3616,1141 "num_input_tokens_seen": 1677721600,1142 "step": 64001143 },1144 {1145 "epoch": 0.04338591773828202,1146 "grad_norm": 0.17044900357723236,1147 "learning_rate": 0.001,1148 "loss": 3.3574,1149 "num_input_tokens_seen": 1690828800,1150 "step": 64501151 },1152 {1153 "epoch": 0.04372224268198963,1154 "grad_norm": 0.18034328520298004,1155 "learning_rate": 0.001,1156 "loss": 3.3625,1157 "num_input_tokens_seen": 1703936000,1158 "step": 65001159 },1160 {1161 "epoch": 0.04372224268198963,1162 "eval_loss": 3.2744550704956055,1163 "eval_runtime": 101.1801,1164 "eval_samples_per_second": 49.417,1165 "eval_steps_per_second": 12.354,1166 "num_input_tokens_seen": 1703936000,1167 "step": 65001168 },1169 {1170 "epoch": 0.044058567625697244,1171 "grad_norm": 0.169066920876503,1172 "learning_rate": 0.001,1173 "loss": 3.3569,1174 "num_input_tokens_seen": 1717043200,1175 "step": 65501176 },1177 {1178 "epoch": 0.044394892569404855,1179 "grad_norm": 0.1789105087518692,1180 "learning_rate": 0.001,1181 "loss": 3.3636,1182 "num_input_tokens_seen": 1730150400,1183 "step": 66001184 },1185 {1186 "epoch": 0.044731217513112466,1187 "grad_norm": 0.2083519697189331,1188 "learning_rate": 0.001,1189 "loss": 3.3565,1190 "num_input_tokens_seen": 1743257600,1191 "step": 66501192 },1193 {1194 "epoch": 0.045067542456820084,1195 "grad_norm": 0.16989745199680328,1196 "learning_rate": 0.001,1197 "loss": 3.3562,1198 "num_input_tokens_seen": 1756364800,1199 "step": 67001200 },