Azrail/smallm_70
072
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 0.333901761331791,6 "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": 0.00023850125809413643,14 "grad_norm": 1.5476292371749878,15 "learning_rate": 0.0001,16 "loss": 9.8285,17 "num_input_tokens_seen": 13107200,18 "step": 5019 },20 {21 "epoch": 0.00047700251618827287,22 "grad_norm": 0.47720426321029663,23 "learning_rate": 0.0002,24 "loss": 7.964,25 "num_input_tokens_seen": 26214400,26 "step": 10027 },28 {29 "epoch": 0.0007155037742824094,30 "grad_norm": 0.8443030714988708,31 "learning_rate": 0.0003,32 "loss": 7.0452,33 "num_input_tokens_seen": 39321600,34 "step": 15035 },36 {37 "epoch": 0.0009540050323765457,38 "grad_norm": 0.5895723104476929,39 "learning_rate": 0.0004,40 "loss": 6.355,41 "num_input_tokens_seen": 52428800,42 "step": 20043 },44 {45 "epoch": 0.0011925062904706823,46 "grad_norm": 0.8343789577484131,47 "learning_rate": 0.0005,48 "loss": 5.8716,49 "num_input_tokens_seen": 65536000,50 "step": 25051 },52 {53 "epoch": 0.0014310075485648188,54 "grad_norm": 0.5747953057289124,55 "learning_rate": 0.0006,56 "loss": 5.5086,57 "num_input_tokens_seen": 78643200,58 "step": 30059 },60 {61 "epoch": 0.0016695088066589552,62 "grad_norm": 0.8383421301841736,63 "learning_rate": 0.0007,64 "loss": 5.2217,65 "num_input_tokens_seen": 91750400,66 "step": 35067 },68 {69 "epoch": 0.0019080100647530915,70 "grad_norm": 0.5696113109588623,71 "learning_rate": 0.0008,72 "loss": 4.9683,73 "num_input_tokens_seen": 104857600,74 "step": 40075 },76 {77 "epoch": 0.002146511322847228,78 "grad_norm": 0.5431691408157349,79 "learning_rate": 0.0009000000000000001,80 "loss": 4.7629,81 "num_input_tokens_seen": 117964800,82 "step": 45083 },84 {85 "epoch": 0.0023850125809413646,86 "grad_norm": 0.4571855664253235,87 "learning_rate": 0.001,88 "loss": 4.5284,89 "num_input_tokens_seen": 131072000,90 "step": 50091 },92 {93 "epoch": 0.0023850125809413646,94 "eval_loss": 4.309167385101318,95 "eval_runtime": 53.3891,96 "eval_samples_per_second": 93.652,97 "eval_steps_per_second": 23.413,98 "num_input_tokens_seen": 131072000,99 "step": 500100 },101 {102 "epoch": 0.002623513839035501,103 "grad_norm": 0.43464773893356323,104 "learning_rate": 0.001,105 "loss": 4.3379,106 "num_input_tokens_seen": 144179200,107 "step": 550108 },109 {110 "epoch": 0.0028620150971296375,111 "grad_norm": 0.49611660838127136,112 "learning_rate": 0.001,113 "loss": 4.1712,114 "num_input_tokens_seen": 157286400,115 "step": 600116 },117 {118 "epoch": 0.0031005163552237738,119 "grad_norm": 0.4060957729816437,120 "learning_rate": 0.001,121 "loss": 4.0436,122 "num_input_tokens_seen": 170393600,123 "step": 650124 },125 {126 "epoch": 0.0033390176133179105,127 "grad_norm": 0.37300577759742737,128 "learning_rate": 0.001,129 "loss": 3.9582,130 "num_input_tokens_seen": 183500800,131 "step": 700132 },133 {134 "epoch": 0.0035775188714120467,135 "grad_norm": 0.4117021858692169,136 "learning_rate": 0.001,137 "loss": 3.8674,138 "num_input_tokens_seen": 196608000,139 "step": 750140 },141 {142 "epoch": 0.003816020129506183,143 "grad_norm": 0.3335980772972107,144 "learning_rate": 0.001,145 "loss": 3.8031,146 "num_input_tokens_seen": 209715200,147 "step": 800148 },149 {150 "epoch": 0.004054521387600319,151 "grad_norm": 0.35943159461021423,152 "learning_rate": 0.001,153 "loss": 3.7534,154 "num_input_tokens_seen": 222822400,155 "step": 850156 },157 {158 "epoch": 0.004293022645694456,159 "grad_norm": 0.40000948309898376,160 "learning_rate": 0.001,161 "loss": 3.6867,162 "num_input_tokens_seen": 235929600,163 "step": 900164 },165 {166 "epoch": 0.0045315239037885926,167 "grad_norm": 0.3165877163410187,168 "learning_rate": 0.001,169 "loss": 3.6565,170 "num_input_tokens_seen": 249036800,171 "step": 950172 },173 {174 "epoch": 0.004770025161882729,175 "grad_norm": 0.3687070906162262,176 "learning_rate": 0.001,177 "loss": 3.6005,178 "num_input_tokens_seen": 262144000,179 "step": 1000180 },181 {182 "epoch": 0.004770025161882729,183 "eval_loss": 3.4853296279907227,184 "eval_runtime": 52.477,185 "eval_samples_per_second": 95.28,186 "eval_steps_per_second": 23.82,187 "num_input_tokens_seen": 262144000,188 "step": 1000189 },190 {191 "epoch": 0.005008526419976865,192 "grad_norm": 0.32389721274375916,193 "learning_rate": 0.001,194 "loss": 3.5663,195 "num_input_tokens_seen": 275251200,196 "step": 1050197 },198 {199 "epoch": 0.005247027678071002,200 "grad_norm": 0.3202049434185028,201 "learning_rate": 0.001,202 "loss": 3.5376,203 "num_input_tokens_seen": 288358400,204 "step": 1100205 },206 {207 "epoch": 0.005485528936165138,208 "grad_norm": 0.30287981033325195,209 "learning_rate": 0.001,210 "loss": 3.5135,211 "num_input_tokens_seen": 301465600,212 "step": 1150213 },214 {215 "epoch": 0.005724030194259275,216 "grad_norm": 0.3624540865421295,217 "learning_rate": 0.001,218 "loss": 3.4814,219 "num_input_tokens_seen": 314572800,220 "step": 1200221 },222 {223 "epoch": 0.005962531452353411,224 "grad_norm": 0.30017992854118347,225 "learning_rate": 0.001,226 "loss": 3.4476,227 "num_input_tokens_seen": 327680000,228 "step": 1250229 },230 {231 "epoch": 0.0062010327104475476,232 "grad_norm": 0.3169330060482025,233 "learning_rate": 0.001,234 "loss": 3.4179,235 "num_input_tokens_seen": 340787200,236 "step": 1300237 },238 {239 "epoch": 0.006439533968541684,240 "grad_norm": 0.2730589210987091,241 "learning_rate": 0.001,242 "loss": 3.4074,243 "num_input_tokens_seen": 353894400,244 "step": 1350245 },246 {247 "epoch": 0.006678035226635821,248 "grad_norm": 0.2927146553993225,249 "learning_rate": 0.001,250 "loss": 3.3757,251 "num_input_tokens_seen": 367001600,252 "step": 1400253 },254 {255 "epoch": 0.006916536484729957,256 "grad_norm": 0.34230080246925354,257 "learning_rate": 0.001,258 "loss": 3.3502,259 "num_input_tokens_seen": 380108800,260 "step": 1450261 },262 {263 "epoch": 0.007155037742824093,264 "grad_norm": 0.30472344160079956,265 "learning_rate": 0.001,266 "loss": 3.3639,267 "num_input_tokens_seen": 393216000,268 "step": 1500269 },270 {271 "epoch": 0.007155037742824093,272 "eval_loss": 3.2486374378204346,273 "eval_runtime": 52.482,274 "eval_samples_per_second": 95.271,275 "eval_steps_per_second": 23.818,276 "num_input_tokens_seen": 393216000,277 "step": 1500278 },279 {280 "epoch": 0.00739353900091823,281 "grad_norm": 0.26124337315559387,282 "learning_rate": 0.001,283 "loss": 3.3521,284 "num_input_tokens_seen": 406323200,285 "step": 1550286 },287 {288 "epoch": 0.007632040259012366,289 "grad_norm": 0.29117754101753235,290 "learning_rate": 0.001,291 "loss": 3.315,292 "num_input_tokens_seen": 419430400,293 "step": 1600294 },295 {296 "epoch": 0.007870541517106503,297 "grad_norm": 0.24080802500247955,298 "learning_rate": 0.001,299 "loss": 3.3103,300 "num_input_tokens_seen": 432537600,301 "step": 1650302 },303 {304 "epoch": 0.008109042775200638,305 "grad_norm": 0.29982003569602966,306 "learning_rate": 0.001,307 "loss": 3.2926,308 "num_input_tokens_seen": 445644800,309 "step": 1700310 },311 {312 "epoch": 0.008347544033294775,313 "grad_norm": 0.26795274019241333,314 "learning_rate": 0.001,315 "loss": 3.2843,316 "num_input_tokens_seen": 458752000,317 "step": 1750318 },319 {320 "epoch": 0.008586045291388912,321 "grad_norm": 0.252774715423584,322 "learning_rate": 0.001,323 "loss": 3.274,324 "num_input_tokens_seen": 471859200,325 "step": 1800326 },327 {328 "epoch": 0.008824546549483048,329 "grad_norm": 0.25432145595550537,330 "learning_rate": 0.001,331 "loss": 3.2533,332 "num_input_tokens_seen": 484966400,333 "step": 1850334 },335 {336 "epoch": 0.009063047807577185,337 "grad_norm": 0.25918108224868774,338 "learning_rate": 0.001,339 "loss": 3.2501,340 "num_input_tokens_seen": 498073600,341 "step": 1900342 },343 {344 "epoch": 0.009301549065671322,345 "grad_norm": 0.2482348382472992,346 "learning_rate": 0.001,347 "loss": 3.2541,348 "num_input_tokens_seen": 511180800,349 "step": 1950350 },351 {352 "epoch": 0.009540050323765458,353 "grad_norm": 0.2615273594856262,354 "learning_rate": 0.001,355 "loss": 3.2218,356 "num_input_tokens_seen": 524288000,357 "step": 2000358 },359 {360 "epoch": 0.009540050323765458,361 "eval_loss": 3.1193039417266846,362 "eval_runtime": 52.7955,363 "eval_samples_per_second": 94.705,364 "eval_steps_per_second": 23.676,365 "num_input_tokens_seen": 524288000,366 "step": 2000367 },368 {369 "epoch": 0.009778551581859595,370 "grad_norm": 0.2637729048728943,371 "learning_rate": 0.001,372 "loss": 3.2285,373 "num_input_tokens_seen": 537395200,374 "step": 2050375 },376 {377 "epoch": 0.01001705283995373,378 "grad_norm": 0.23936080932617188,379 "learning_rate": 0.001,380 "loss": 3.2119,381 "num_input_tokens_seen": 550502400,382 "step": 2100383 },384 {385 "epoch": 0.010255554098047867,386 "grad_norm": 0.2469020038843155,387 "learning_rate": 0.001,388 "loss": 3.2021,389 "num_input_tokens_seen": 563609600,390 "step": 2150391 },392 {393 "epoch": 0.010494055356142003,394 "grad_norm": 0.2304004430770874,395 "learning_rate": 0.001,396 "loss": 3.1874,397 "num_input_tokens_seen": 576716800,398 "step": 2200399 },400 {401 "epoch": 0.01073255661423614,402 "grad_norm": 0.232864648103714,403 "learning_rate": 0.001,404 "loss": 3.1897,405 "num_input_tokens_seen": 589824000,406 "step": 2250407 },408 {409 "epoch": 0.010971057872330277,410 "grad_norm": 0.23161470890045166,411 "learning_rate": 0.001,412 "loss": 3.1689,413 "num_input_tokens_seen": 602931200,414 "step": 2300415 },416 {417 "epoch": 0.011209559130424413,418 "grad_norm": 0.20868408679962158,419 "learning_rate": 0.001,420 "loss": 3.1615,421 "num_input_tokens_seen": 616038400,422 "step": 2350423 },424 {425 "epoch": 0.01144806038851855,426 "grad_norm": 0.23374608159065247,427 "learning_rate": 0.001,428 "loss": 3.1556,429 "num_input_tokens_seen": 629145600,430 "step": 2400431 },432 {433 "epoch": 0.011686561646612685,434 "grad_norm": 0.21716611087322235,435 "learning_rate": 0.001,436 "loss": 3.1463,437 "num_input_tokens_seen": 642252800,438 "step": 2450439 },440 {441 "epoch": 0.011925062904706822,442 "grad_norm": 0.23689670860767365,443 "learning_rate": 0.001,444 "loss": 3.1433,445 "num_input_tokens_seen": 655360000,446 "step": 2500447 },448 {449 "epoch": 0.011925062904706822,450 "eval_loss": 3.040046215057373,451 "eval_runtime": 52.9109,452 "eval_samples_per_second": 94.498,453 "eval_steps_per_second": 23.625,454 "num_input_tokens_seen": 655360000,455 "step": 2500456 },457 {458 "epoch": 0.012163564162800958,459 "grad_norm": 0.2245575189590454,460 "learning_rate": 0.001,461 "loss": 3.1445,462 "num_input_tokens_seen": 668467200,463 "step": 2550464 },465 {466 "epoch": 0.012402065420895095,467 "grad_norm": 0.20992259681224823,468 "learning_rate": 0.001,469 "loss": 3.1447,470 "num_input_tokens_seen": 681574400,471 "step": 2600472 },473 {474 "epoch": 0.012640566678989232,475 "grad_norm": 0.21792201697826385,476 "learning_rate": 0.001,477 "loss": 3.1323,478 "num_input_tokens_seen": 694681600,479 "step": 2650480 },481 {482 "epoch": 0.012879067937083368,483 "grad_norm": 0.243458554148674,484 "learning_rate": 0.001,485 "loss": 3.1084,486 "num_input_tokens_seen": 707788800,487 "step": 2700488 },489 {490 "epoch": 0.013117569195177505,491 "grad_norm": 0.21190515160560608,492 "learning_rate": 0.001,493 "loss": 3.1202,494 "num_input_tokens_seen": 720896000,495 "step": 2750496 },497 {498 "epoch": 0.013356070453271642,499 "grad_norm": 0.2461613118648529,500 "learning_rate": 0.001,501 "loss": 3.1007,502 "num_input_tokens_seen": 734003200,503 "step": 2800504 },505 {506 "epoch": 0.013594571711365777,507 "grad_norm": 0.1976248323917389,508 "learning_rate": 0.001,509 "loss": 3.1079,510 "num_input_tokens_seen": 747110400,511 "step": 2850512 },513 {514 "epoch": 0.013833072969459913,515 "grad_norm": 0.22097842395305634,516 "learning_rate": 0.001,517 "loss": 3.0846,518 "num_input_tokens_seen": 760217600,519 "step": 2900520 },521 {522 "epoch": 0.01407157422755405,523 "grad_norm": 0.20581132173538208,524 "learning_rate": 0.001,525 "loss": 3.0995,526 "num_input_tokens_seen": 773324800,527 "step": 2950528 },529 {530 "epoch": 0.014310075485648187,531 "grad_norm": 0.19790051877498627,532 "learning_rate": 0.001,533 "loss": 3.0977,534 "num_input_tokens_seen": 786432000,535 "step": 3000536 },537 {538 "epoch": 0.014310075485648187,539 "eval_loss": 2.9804909229278564,540 "eval_runtime": 53.1278,541 "eval_samples_per_second": 94.113,542 "eval_steps_per_second": 23.528,543 "num_input_tokens_seen": 786432000,544 "step": 3000545 },546 {547 "epoch": 0.014548576743742323,548 "grad_norm": 0.20328116416931152,549 "learning_rate": 0.001,550 "loss": 3.0872,551 "num_input_tokens_seen": 799539200,552 "step": 3050553 },554 {555 "epoch": 0.01478707800183646,556 "grad_norm": 0.21318025887012482,557 "learning_rate": 0.001,558 "loss": 3.0861,559 "num_input_tokens_seen": 812646400,560 "step": 3100561 },562 {563 "epoch": 0.015025579259930597,564 "grad_norm": 0.22170069813728333,565 "learning_rate": 0.001,566 "loss": 3.0618,567 "num_input_tokens_seen": 825753600,568 "step": 3150569 },570 {571 "epoch": 0.015264080518024732,572 "grad_norm": 0.21292312443256378,573 "learning_rate": 0.001,574 "loss": 3.0567,575 "num_input_tokens_seen": 838860800,576 "step": 3200577 },578 {579 "epoch": 0.015502581776118868,580 "grad_norm": 0.2331959754228592,581 "learning_rate": 0.001,582 "loss": 3.0714,583 "num_input_tokens_seen": 851968000,584 "step": 3250585 },586 {587 "epoch": 0.015741083034213007,588 "grad_norm": 0.19236011803150177,589 "learning_rate": 0.001,590 "loss": 3.059,591 "num_input_tokens_seen": 865075200,592 "step": 3300593 },594 {595 "epoch": 0.015979584292307142,596 "grad_norm": 0.19991376996040344,597 "learning_rate": 0.001,598 "loss": 3.0542,599 "num_input_tokens_seen": 878182400,600 "step": 3350601 },602 {603 "epoch": 0.016218085550401277,604 "grad_norm": 0.2042934149503708,605 "learning_rate": 0.001,606 "loss": 3.0517,607 "num_input_tokens_seen": 891289600,608 "step": 3400609 },610 {611 "epoch": 0.016456586808495415,612 "grad_norm": 0.19254428148269653,613 "learning_rate": 0.001,614 "loss": 3.0415,615 "num_input_tokens_seen": 904396800,616 "step": 3450617 },618 {619 "epoch": 0.01669508806658955,620 "grad_norm": 0.19211998581886292,621 "learning_rate": 0.001,622 "loss": 3.0253,623 "num_input_tokens_seen": 917504000,624 "step": 3500625 },626 {627 "epoch": 0.01669508806658955,628 "eval_loss": 2.937037944793701,629 "eval_runtime": 52.6773,630 "eval_samples_per_second": 94.918,631 "eval_steps_per_second": 23.729,632 "num_input_tokens_seen": 917504000,633 "step": 3500634 },635 {636 "epoch": 0.01693358932468369,637 "grad_norm": 0.19596482813358307,638 "learning_rate": 0.001,639 "loss": 3.053,640 "num_input_tokens_seen": 930611200,641 "step": 3550642 },643 {644 "epoch": 0.017172090582777823,645 "grad_norm": 0.20214103162288666,646 "learning_rate": 0.001,647 "loss": 3.0385,648 "num_input_tokens_seen": 943718400,649 "step": 3600650 },651 {652 "epoch": 0.017410591840871962,653 "grad_norm": 0.18580283224582672,654 "learning_rate": 0.001,655 "loss": 3.0354,656 "num_input_tokens_seen": 956825600,657 "step": 3650658 },659 {660 "epoch": 0.017649093098966097,661 "grad_norm": 0.18928515911102295,662 "learning_rate": 0.001,663 "loss": 3.0292,664 "num_input_tokens_seen": 969932800,665 "step": 3700666 },667 {668 "epoch": 0.017887594357060232,669 "grad_norm": 0.19066137075424194,670 "learning_rate": 0.001,671 "loss": 3.0206,672 "num_input_tokens_seen": 983040000,673 "step": 3750674 },675 {676 "epoch": 0.01812609561515437,677 "grad_norm": 0.20291416347026825,678 "learning_rate": 0.001,679 "loss": 3.0254,680 "num_input_tokens_seen": 996147200,681 "step": 3800682 },683 {684 "epoch": 0.018364596873248505,685 "grad_norm": 0.19991491734981537,686 "learning_rate": 0.001,687 "loss": 3.0212,688 "num_input_tokens_seen": 1009254400,689 "step": 3850690 },691 {692 "epoch": 0.018603098131342644,693 "grad_norm": 0.19553051888942719,694 "learning_rate": 0.001,695 "loss": 3.0229,696 "num_input_tokens_seen": 1022361600,697 "step": 3900698 },699 {700 "epoch": 0.01884159938943678,701 "grad_norm": 0.19302095472812653,702 "learning_rate": 0.001,703 "loss": 3.0137,704 "num_input_tokens_seen": 1035468800,705 "step": 3950706 },707 {708 "epoch": 0.019080100647530917,709 "grad_norm": 0.18680201470851898,710 "learning_rate": 0.001,711 "loss": 3.0106,712 "num_input_tokens_seen": 1048576000,713 "step": 4000714 },715 {716 "epoch": 0.019080100647530917,717 "eval_loss": 2.8984477519989014,718 "eval_runtime": 52.851,719 "eval_samples_per_second": 94.606,720 "eval_steps_per_second": 23.651,721 "num_input_tokens_seen": 1048576000,722 "step": 4000723 },724 {725 "epoch": 0.019318601905625052,726 "grad_norm": 0.18222174048423767,727 "learning_rate": 0.001,728 "loss": 3.0095,729 "num_input_tokens_seen": 1061683200,730 "step": 4050731 },732 {733 "epoch": 0.01955710316371919,734 "grad_norm": 0.1929137110710144,735 "learning_rate": 0.001,736 "loss": 3.0022,737 "num_input_tokens_seen": 1074790400,738 "step": 4100739 },740 {741 "epoch": 0.019795604421813325,742 "grad_norm": 0.19358602166175842,743 "learning_rate": 0.001,744 "loss": 2.9978,745 "num_input_tokens_seen": 1087897600,746 "step": 4150747 },748 {749 "epoch": 0.02003410567990746,750 "grad_norm": 0.19070614874362946,751 "learning_rate": 0.001,752 "loss": 3.0016,753 "num_input_tokens_seen": 1101004800,754 "step": 4200755 },756 {757 "epoch": 0.0202726069380016,758 "grad_norm": 0.17888160049915314,759 "learning_rate": 0.001,760 "loss": 2.9984,761 "num_input_tokens_seen": 1114112000,762 "step": 4250763 },764 {765 "epoch": 0.020511108196095734,766 "grad_norm": 0.1823708564043045,767 "learning_rate": 0.001,768 "loss": 3.004,769 "num_input_tokens_seen": 1127219200,770 "step": 4300771 },772 {773 "epoch": 0.020749609454189872,774 "grad_norm": 0.1753600388765335,775 "learning_rate": 0.001,776 "loss": 2.9814,777 "num_input_tokens_seen": 1140326400,778 "step": 4350779 },780 {781 "epoch": 0.020988110712284007,782 "grad_norm": 0.1710510551929474,783 "learning_rate": 0.001,784 "loss": 2.9597,785 "num_input_tokens_seen": 1153433600,786 "step": 4400787 },788 {789 "epoch": 0.021226611970378145,790 "grad_norm": 0.18727277219295502,791 "learning_rate": 0.001,792 "loss": 2.9695,793 "num_input_tokens_seen": 1166540800,794 "step": 4450795 },796 {797 "epoch": 0.02146511322847228,798 "grad_norm": 0.17773132026195526,799 "learning_rate": 0.001,800 "loss": 2.9664,801 "num_input_tokens_seen": 1179648000,802 "step": 4500803 },804 {805 "epoch": 0.02146511322847228,806 "eval_loss": 2.871137857437134,807 "eval_runtime": 51.4876,808 "eval_samples_per_second": 97.111,809 "eval_steps_per_second": 24.278,810 "num_input_tokens_seen": 1179648000,811 "step": 4500812 },813 {814 "epoch": 0.021703614486566415,815 "grad_norm": 0.1875799000263214,816 "learning_rate": 0.001,817 "loss": 2.9682,818 "num_input_tokens_seen": 1192755200,819 "step": 4550820 },821 {822 "epoch": 0.021942115744660554,823 "grad_norm": 0.18222226202487946,824 "learning_rate": 0.001,825 "loss": 2.9484,826 "num_input_tokens_seen": 1205862400,827 "step": 4600828 },829 {830 "epoch": 0.02218061700275469,831 "grad_norm": 0.191411092877388,832 "learning_rate": 0.001,833 "loss": 2.9637,834 "num_input_tokens_seen": 1218969600,835 "step": 4650836 },837 {838 "epoch": 0.022419118260848827,839 "grad_norm": 0.17608201503753662,840 "learning_rate": 0.001,841 "loss": 2.9792,842 "num_input_tokens_seen": 1232076800,843 "step": 4700844 },845 {846 "epoch": 0.022657619518942962,847 "grad_norm": 0.1718858927488327,848 "learning_rate": 0.001,849 "loss": 2.9674,850 "num_input_tokens_seen": 1245184000,851 "step": 4750852 },853 {854 "epoch": 0.0228961207770371,855 "grad_norm": 0.18428942561149597,856 "learning_rate": 0.001,857 "loss": 2.976,858 "num_input_tokens_seen": 1258291200,859 "step": 4800860 },861 {862 "epoch": 0.023134622035131235,863 "grad_norm": 0.16696259379386902,864 "learning_rate": 0.001,865 "loss": 2.9486,866 "num_input_tokens_seen": 1271398400,867 "step": 4850868 },869 {870 "epoch": 0.02337312329322537,871 "grad_norm": 0.18239040672779083,872 "learning_rate": 0.001,873 "loss": 2.956,874 "num_input_tokens_seen": 1284505600,875 "step": 4900876 },877 {878 "epoch": 0.02361162455131951,879 "grad_norm": 0.17167994379997253,880 "learning_rate": 0.001,881 "loss": 2.9449,882 "num_input_tokens_seen": 1297612800,883 "step": 4950884 },885 {886 "epoch": 0.023850125809413644,887 "grad_norm": 0.18532761931419373,888 "learning_rate": 0.001,889 "loss": 2.947,890 "num_input_tokens_seen": 1310720000,891 "step": 5000892 },893 {894 "epoch": 0.023850125809413644,895 "eval_loss": 2.8470754623413086,896 "eval_runtime": 51.04,897 "eval_samples_per_second": 97.962,898 "eval_steps_per_second": 24.491,899 "num_input_tokens_seen": 1310720000,900 "step": 5000901 },902 {903 "epoch": 0.024088627067507782,904 "grad_norm": 0.21697266399860382,905 "learning_rate": 0.001,906 "loss": 2.963,907 "num_input_tokens_seen": 1323827200,908 "step": 5050909 },910 {911 "epoch": 0.024327128325601917,912 "grad_norm": 0.17018833756446838,913 "learning_rate": 0.001,914 "loss": 2.9453,915 "num_input_tokens_seen": 1336934400,916 "step": 5100917 },918 {919 "epoch": 0.024565629583696055,920 "grad_norm": 0.17473167181015015,921 "learning_rate": 0.001,922 "loss": 2.9516,923 "num_input_tokens_seen": 1350041600,924 "step": 5150925 },926 {927 "epoch": 0.02480413084179019,928 "grad_norm": 0.18488293886184692,929 "learning_rate": 0.001,930 "loss": 2.9404,931 "num_input_tokens_seen": 1363148800,932 "step": 5200933 },934 {935 "epoch": 0.025042632099884325,936 "grad_norm": 0.17348967492580414,937 "learning_rate": 0.001,938 "loss": 2.9275,939 "num_input_tokens_seen": 1376256000,940 "step": 5250941 },942 {943 "epoch": 0.025281133357978464,944 "grad_norm": 0.16547563672065735,945 "learning_rate": 0.001,946 "loss": 2.9464,947 "num_input_tokens_seen": 1389363200,948 "step": 5300949 },950 {951 "epoch": 0.0255196346160726,952 "grad_norm": 0.17538361251354218,953 "learning_rate": 0.001,954 "loss": 2.94,955 "num_input_tokens_seen": 1402470400,956 "step": 5350957 },958 {959 "epoch": 0.025758135874166737,960 "grad_norm": 0.17068558931350708,961 "learning_rate": 0.001,962 "loss": 2.9382,963 "num_input_tokens_seen": 1415577600,964 "step": 5400965 },966 {967 "epoch": 0.025996637132260872,968 "grad_norm": 0.17389337718486786,969 "learning_rate": 0.001,970 "loss": 2.9254,971 "num_input_tokens_seen": 1428684800,972 "step": 5450973 },974 {975 "epoch": 0.02623513839035501,976 "grad_norm": 0.17620491981506348,977 "learning_rate": 0.001,978 "loss": 2.9221,979 "num_input_tokens_seen": 1441792000,980 "step": 5500981 },982 {983 "epoch": 0.02623513839035501,984 "eval_loss": 2.8246922492980957,985 "eval_runtime": 50.2832,986 "eval_samples_per_second": 99.437,987 "eval_steps_per_second": 24.859,988 "num_input_tokens_seen": 1441792000,989 "step": 5500990 },991 {992 "epoch": 0.026473639648449145,993 "grad_norm": 0.15889622271060944,994 "learning_rate": 0.001,995 "loss": 2.923,996 "num_input_tokens_seen": 1454899200,997 "step": 5550998 },999 {1000 "epoch": 0.026712140906543284,1001 "grad_norm": 0.17490123212337494,1002 "learning_rate": 0.001,1003 "loss": 2.9146,1004 "num_input_tokens_seen": 1468006400,1005 "step": 56001006 },1007 {1008 "epoch": 0.02695064216463742,1009 "grad_norm": 0.17789559066295624,1010 "learning_rate": 0.001,1011 "loss": 2.9253,1012 "num_input_tokens_seen": 1481113600,1013 "step": 56501014 },1015 {1016 "epoch": 0.027189143422731554,1017 "grad_norm": 0.17113780975341797,1018 "learning_rate": 0.001,1019 "loss": 2.9267,1020 "num_input_tokens_seen": 1494220800,1021 "step": 57001022 },1023 {1024 "epoch": 0.027427644680825692,1025 "grad_norm": 0.1671907901763916,1026 "learning_rate": 0.001,1027 "loss": 2.9178,1028 "num_input_tokens_seen": 1507328000,1029 "step": 57501030 },1031 {1032 "epoch": 0.027666145938919827,1033 "grad_norm": 0.17511603236198425,1034 "learning_rate": 0.001,1035 "loss": 2.9341,1036 "num_input_tokens_seen": 1520435200,1037 "step": 58001038 },1039 {1040 "epoch": 0.027904647197013965,1041 "grad_norm": 0.1821524053812027,1042 "learning_rate": 0.001,1043 "loss": 2.9076,1044 "num_input_tokens_seen": 1533542400,1045 "step": 58501046 },1047 {1048 "epoch": 0.0281431484551081,1049 "grad_norm": 0.16259051859378815,1050 "learning_rate": 0.001,1051 "loss": 2.9212,1052 "num_input_tokens_seen": 1546649600,1053 "step": 59001054 },1055 {1056 "epoch": 0.02838164971320224,1057 "grad_norm": 0.18584352731704712,1058 "learning_rate": 0.001,1059 "loss": 2.927,1060 "num_input_tokens_seen": 1559756800,1061 "step": 59501062 },1063 {1064 "epoch": 0.028620150971296374,1065 "grad_norm": 0.181602343916893,1066 "learning_rate": 0.001,1067 "loss": 2.9096,1068 "num_input_tokens_seen": 1572864000,1069 "step": 60001070 },1071 {1072 "epoch": 0.028620150971296374,1073 "eval_loss": 2.8036017417907715,1074 "eval_runtime": 50.543,1075 "eval_samples_per_second": 98.926,1076 "eval_steps_per_second": 24.731,1077 "num_input_tokens_seen": 1572864000,1078 "step": 60001079 },1080 {1081 "epoch": 0.02885865222939051,1082 "grad_norm": 0.1653270423412323,1083 "learning_rate": 0.001,1084 "loss": 2.9214,1085 "num_input_tokens_seen": 1585971200,1086 "step": 60501087 },1088 {1089 "epoch": 0.029097153487484647,1090 "grad_norm": 0.17030183970928192,1091 "learning_rate": 0.001,1092 "loss": 2.9081,1093 "num_input_tokens_seen": 1599078400,1094 "step": 61001095 },1096 {1097 "epoch": 0.029335654745578782,1098 "grad_norm": 0.17734774947166443,1099 "learning_rate": 0.001,1100 "loss": 2.9128,1101 "num_input_tokens_seen": 1612185600,1102 "step": 61501103 },1104 {1105 "epoch": 0.02957415600367292,1106 "grad_norm": 0.1664343774318695,1107 "learning_rate": 0.001,1108 "loss": 2.9084,1109 "num_input_tokens_seen": 1625292800,1110 "step": 62001111 },1112 {1113 "epoch": 0.029812657261767055,1114 "grad_norm": 0.15939603745937347,1115 "learning_rate": 0.001,1116 "loss": 2.9049,1117 "num_input_tokens_seen": 1638400000,1118 "step": 62501119 },1120 {1121 "epoch": 0.030051158519861194,1122 "grad_norm": 0.16107864677906036,1123 "learning_rate": 0.001,1124 "loss": 2.8889,1125 "num_input_tokens_seen": 1651507200,1126 "step": 63001127 },1128 {1129 "epoch": 0.03028965977795533,1130 "grad_norm": 0.1734771579504013,1131 "learning_rate": 0.001,1132 "loss": 2.8951,1133 "num_input_tokens_seen": 1664614400,1134 "step": 63501135 },1136 {1137 "epoch": 0.030528161036049464,1138 "grad_norm": 0.1804204136133194,1139 "learning_rate": 0.001,1140 "loss": 2.8877,1141 "num_input_tokens_seen": 1677721600,1142 "step": 64001143 },1144 {1145 "epoch": 0.030766662294143602,1146 "grad_norm": 0.16369500756263733,1147 "learning_rate": 0.001,1148 "loss": 2.8764,1149 "num_input_tokens_seen": 1690828800,1150 "step": 64501151 },1152 {1153 "epoch": 0.031005163552237737,1154 "grad_norm": 0.1704144924879074,1155 "learning_rate": 0.001,1156 "loss": 2.8965,1157 "num_input_tokens_seen": 1703936000,1158 "step": 65001159 },1160 {1161 "epoch": 0.031005163552237737,1162 "eval_loss": 2.787343740463257,1163 "eval_runtime": 50.3956,1164 "eval_samples_per_second": 99.215,1165 "eval_steps_per_second": 24.804,1166 "num_input_tokens_seen": 1703936000,1167 "step": 65001168 },1169 {1170 "epoch": 0.031243664810331875,1171 "grad_norm": 0.17917555570602417,1172 "learning_rate": 0.001,1173 "loss": 2.8882,1174 "num_input_tokens_seen": 1717043200,1175 "step": 65501176 },1177 {1178 "epoch": 0.031482166068426014,1179 "grad_norm": 0.18822412192821503,1180 "learning_rate": 0.001,1181 "loss": 2.8931,1182 "num_input_tokens_seen": 1730150400,1183 "step": 66001184 },1185 {1186 "epoch": 0.031720667326520145,1187 "grad_norm": 0.1702752560377121,1188 "learning_rate": 0.001,1189 "loss": 2.8906,1190 "num_input_tokens_seen": 1743257600,1191 "step": 66501192 },1193 {1194 "epoch": 0.031959168584614284,1195 "grad_norm": 0.16963082551956177,1196 "learning_rate": 0.001,1197 "loss": 2.8809,1198 "num_input_tokens_seen": 1756364800,1199 "step": 67001200 },