gtonkov/SpecAI
016
1{2 "best_metric": null,3 "best_model_checkpoint": null,4 "epoch": 20.217435897435898,5 "eval_steps": 500,6 "global_step": 4928,7 "is_hyper_param_search": false,8 "is_local_process_zero": true,9 "is_world_process_zero": true,10 "log_history": [11 {12 "epoch": 0.020512820512820513,13 "grad_norm": 0.875,14 "learning_rate": 2.9999969395283144e-06,15 "loss": 0.6637,16 "num_input_tokens_seen": 75456,17 "step": 518 },19 {20 "epoch": 0.041025641025641026,21 "grad_norm": 1.4140625,22 "learning_rate": 2.9999877581257458e-06,23 "loss": 0.6963,24 "num_input_tokens_seen": 160992,25 "step": 1026 },27 {28 "epoch": 0.06153846153846154,29 "grad_norm": 1.59375,30 "learning_rate": 2.9999724558297605e-06,31 "loss": 0.6919,32 "num_input_tokens_seen": 253856,33 "step": 1534 },35 {36 "epoch": 0.08205128205128205,37 "grad_norm": 2.40625,38 "learning_rate": 2.999951032702801e-06,39 "loss": 0.6605,40 "num_input_tokens_seen": 328640,41 "step": 2042 },43 {44 "epoch": 0.10256410256410256,45 "grad_norm": 0.75390625,46 "learning_rate": 2.9999234888322877e-06,47 "loss": 0.5888,48 "num_input_tokens_seen": 401504,49 "step": 2550 },51 {52 "epoch": 0.12307692307692308,53 "grad_norm": 1.578125,54 "learning_rate": 2.9998898243306162e-06,55 "loss": 0.6087,56 "num_input_tokens_seen": 475168,57 "step": 3058 },59 {60 "epoch": 0.14358974358974358,61 "grad_norm": 0.89453125,62 "learning_rate": 2.99985003933516e-06,63 "loss": 0.6405,64 "num_input_tokens_seen": 547168,65 "step": 3566 },67 {68 "epoch": 0.1641025641025641,69 "grad_norm": 1.1015625,70 "learning_rate": 2.999804134008266e-06,71 "loss": 0.6443,72 "num_input_tokens_seen": 618176,73 "step": 4074 },75 {76 "epoch": 0.18461538461538463,77 "grad_norm": 0.69921875,78 "learning_rate": 2.9997521085372565e-06,79 "loss": 0.6554,80 "num_input_tokens_seen": 701184,81 "step": 4582 },83 {84 "epoch": 0.20512820512820512,85 "grad_norm": 1.0234375,86 "learning_rate": 2.999693963134429e-06,87 "loss": 0.6903,88 "num_input_tokens_seen": 782656,89 "step": 5090 },91 {92 "epoch": 0.22564102564102564,93 "grad_norm": 0.796875,94 "learning_rate": 2.9996296980370526e-06,95 "loss": 0.6915,96 "num_input_tokens_seen": 869280,97 "step": 5598 },99 {100 "epoch": 0.24615384615384617,101 "grad_norm": 0.8515625,102 "learning_rate": 2.99955931350737e-06,103 "loss": 0.584,104 "num_input_tokens_seen": 944032,105 "step": 60106 },107 {108 "epoch": 0.26666666666666666,109 "grad_norm": 0.91796875,110 "learning_rate": 2.999482809832594e-06,111 "loss": 0.5908,112 "num_input_tokens_seen": 1025568,113 "step": 65114 },115 {116 "epoch": 0.28717948717948716,117 "grad_norm": 0.84375,118 "learning_rate": 2.9994001873249074e-06,119 "loss": 0.5811,120 "num_input_tokens_seen": 1105344,121 "step": 70122 },123 {124 "epoch": 0.3076923076923077,125 "grad_norm": 0.6953125,126 "learning_rate": 2.999311446321462e-06,127 "loss": 0.5792,128 "num_input_tokens_seen": 1182016,129 "step": 75130 },131 {132 "epoch": 0.3282051282051282,133 "grad_norm": 0.99609375,134 "learning_rate": 2.999216587184378e-06,135 "loss": 0.5923,136 "num_input_tokens_seen": 1265440,137 "step": 80138 },139 {140 "epoch": 0.3487179487179487,141 "grad_norm": 0.91015625,142 "learning_rate": 2.9991156103007394e-06,143 "loss": 0.5896,144 "num_input_tokens_seen": 1347296,145 "step": 85146 },147 {148 "epoch": 0.36923076923076925,149 "grad_norm": 0.90234375,150 "learning_rate": 2.9990085160825954e-06,151 "loss": 0.6283,152 "num_input_tokens_seen": 1424064,153 "step": 90154 },155 {156 "epoch": 0.38974358974358975,157 "grad_norm": 0.67578125,158 "learning_rate": 2.9988953049669577e-06,159 "loss": 0.5329,160 "num_input_tokens_seen": 1492416,161 "step": 95162 },163 {164 "epoch": 0.41025641025641024,165 "grad_norm": 0.44921875,166 "learning_rate": 2.998775977415799e-06,167 "loss": 0.5582,168 "num_input_tokens_seen": 1572192,169 "step": 100170 },171 {172 "epoch": 0.4307692307692308,173 "grad_norm": 0.59375,174 "learning_rate": 2.998650533916051e-06,175 "loss": 0.6529,176 "num_input_tokens_seen": 1650592,177 "step": 105178 },179 {180 "epoch": 0.4512820512820513,181 "grad_norm": 0.515625,182 "learning_rate": 2.998518974979602e-06,183 "loss": 0.5449,184 "num_input_tokens_seen": 1728128,185 "step": 110186 },187 {188 "epoch": 0.4717948717948718,189 "grad_norm": 0.494140625,190 "learning_rate": 2.998381301143295e-06,191 "loss": 0.573,192 "num_input_tokens_seen": 1806080,193 "step": 115194 },195 {196 "epoch": 0.49230769230769234,197 "grad_norm": 0.38671875,198 "learning_rate": 2.9982375129689253e-06,199 "loss": 0.6065,200 "num_input_tokens_seen": 1884544,201 "step": 120202 },203 {204 "epoch": 0.5128205128205128,205 "grad_norm": 0.396484375,206 "learning_rate": 2.9980876110432404e-06,207 "loss": 0.6232,208 "num_input_tokens_seen": 1961376,209 "step": 125210 },211 {212 "epoch": 0.5333333333333333,213 "grad_norm": 0.408203125,214 "learning_rate": 2.9979315959779335e-06,215 "loss": 0.5061,216 "num_input_tokens_seen": 2036064,217 "step": 130218 },219 {220 "epoch": 0.5538461538461539,221 "grad_norm": 0.271484375,222 "learning_rate": 2.9977694684096447e-06,223 "loss": 0.5657,224 "num_input_tokens_seen": 2109376,225 "step": 135226 },227 {228 "epoch": 0.5743589743589743,229 "grad_norm": 0.27734375,230 "learning_rate": 2.997601228999956e-06,231 "loss": 0.5597,232 "num_input_tokens_seen": 2181344,233 "step": 140234 },235 {236 "epoch": 0.5948717948717949,237 "grad_norm": 0.2451171875,238 "learning_rate": 2.99742687843539e-06,239 "loss": 0.5514,240 "num_input_tokens_seen": 2256768,241 "step": 145242 },243 {244 "epoch": 0.6153846153846154,245 "grad_norm": 0.2265625,246 "learning_rate": 2.997246417427407e-06,247 "loss": 0.53,248 "num_input_tokens_seen": 2330144,249 "step": 150250 },251 {252 "epoch": 0.6358974358974359,253 "grad_norm": 0.19921875,254 "learning_rate": 2.9970598467124008e-06,255 "loss": 0.5365,256 "num_input_tokens_seen": 2402688,257 "step": 155258 },259 {260 "epoch": 0.6564102564102564,261 "grad_norm": 0.1923828125,262 "learning_rate": 2.9968671670516983e-06,263 "loss": 0.5578,264 "num_input_tokens_seen": 2476800,265 "step": 160266 },267 {268 "epoch": 0.676923076923077,269 "grad_norm": 0.166015625,270 "learning_rate": 2.9966683792315528e-06,271 "loss": 0.5492,272 "num_input_tokens_seen": 2565792,273 "step": 165274 },275 {276 "epoch": 0.6974358974358974,277 "grad_norm": 0.1533203125,278 "learning_rate": 2.9964634840631435e-06,279 "loss": 0.5144,280 "num_input_tokens_seen": 2637792,281 "step": 170282 },283 {284 "epoch": 0.717948717948718,285 "grad_norm": 0.154296875,286 "learning_rate": 2.9962524823825724e-06,287 "loss": 0.5741,288 "num_input_tokens_seen": 2718944,289 "step": 175290 },291 {292 "epoch": 0.7384615384615385,293 "grad_norm": 0.1513671875,294 "learning_rate": 2.9960353750508583e-06,295 "loss": 0.5846,296 "num_input_tokens_seen": 2802240,297 "step": 180298 },299 {300 "epoch": 0.7589743589743589,301 "grad_norm": 0.1474609375,302 "learning_rate": 2.995812162953936e-06,303 "loss": 0.5834,304 "num_input_tokens_seen": 2884672,305 "step": 185306 },307 {308 "epoch": 0.7794871794871795,309 "grad_norm": 0.154296875,310 "learning_rate": 2.9955828470026515e-06,311 "loss": 0.5493,312 "num_input_tokens_seen": 2953536,313 "step": 190314 },315 {316 "epoch": 0.8,317 "grad_norm": 0.138671875,318 "learning_rate": 2.9953474281327576e-06,319 "loss": 0.5503,320 "num_input_tokens_seen": 3026496,321 "step": 195322 },323 {324 "epoch": 0.8205128205128205,325 "grad_norm": 0.244140625,326 "learning_rate": 2.995105907304912e-06,327 "loss": 0.5835,328 "num_input_tokens_seen": 3105376,329 "step": 200330 },331 {332 "epoch": 0.841025641025641,333 "grad_norm": 0.349609375,334 "learning_rate": 2.9948582855046704e-06,335 "loss": 0.5583,336 "num_input_tokens_seen": 3179776,337 "step": 205338 },339 {340 "epoch": 0.8615384615384616,341 "grad_norm": 0.1826171875,342 "learning_rate": 2.9946045637424864e-06,343 "loss": 0.6392,344 "num_input_tokens_seen": 3262336,345 "step": 210346 },347 {348 "epoch": 0.882051282051282,349 "grad_norm": 0.1640625,350 "learning_rate": 2.994344743053704e-06,351 "loss": 0.5658,352 "num_input_tokens_seen": 3343904,353 "step": 215354 },355 {356 "epoch": 0.9025641025641026,357 "grad_norm": 0.50390625,358 "learning_rate": 2.9940788244985557e-06,359 "loss": 0.5609,360 "num_input_tokens_seen": 3419264,361 "step": 220362 },363 {364 "epoch": 0.9230769230769231,365 "grad_norm": 0.458984375,366 "learning_rate": 2.9938068091621556e-06,367 "loss": 0.5305,368 "num_input_tokens_seen": 3496032,369 "step": 225370 },371 {372 "epoch": 0.9435897435897436,373 "grad_norm": 0.41015625,374 "learning_rate": 2.9935286981544975e-06,375 "loss": 0.5652,376 "num_input_tokens_seen": 3570880,377 "step": 230378 },379 {380 "epoch": 0.9641025641025641,381 "grad_norm": 0.7734375,382 "learning_rate": 2.9932444926104495e-06,383 "loss": 0.5348,384 "num_input_tokens_seen": 3644096,385 "step": 235386 },387 {388 "epoch": 0.9846153846153847,389 "grad_norm": 0.86328125,390 "learning_rate": 2.992954193689749e-06,391 "loss": 0.5344,392 "num_input_tokens_seen": 3717088,393 "step": 240394 },395 {396 "epoch": 1.005128205128205,397 "grad_norm": 0.435546875,398 "learning_rate": 2.9926578025769978e-06,399 "loss": 0.5977,400 "num_input_tokens_seen": 3783648,401 "step": 245402 },403 {404 "epoch": 1.0256410256410255,405 "grad_norm": 0.408203125,406 "learning_rate": 2.992355320481658e-06,407 "loss": 0.5304,408 "num_input_tokens_seen": 3852160,409 "step": 250410 },411 {412 "epoch": 1.0461538461538462,413 "grad_norm": 0.75390625,414 "learning_rate": 2.9920467486380475e-06,415 "loss": 0.5617,416 "num_input_tokens_seen": 3924416,417 "step": 255418 },419 {420 "epoch": 1.0666666666666667,421 "grad_norm": 0.46875,422 "learning_rate": 2.991732088305333e-06,423 "loss": 0.5367,424 "num_input_tokens_seen": 4006432,425 "step": 260426 },427 {428 "epoch": 1.087179487179487,429 "grad_norm": 0.5234375,430 "learning_rate": 2.991411340767526e-06,431 "loss": 0.5207,432 "num_input_tokens_seen": 4081888,433 "step": 265434 },435 {436 "epoch": 1.1076923076923078,437 "grad_norm": 0.41015625,438 "learning_rate": 2.9910845073334793e-06,439 "loss": 0.556,440 "num_input_tokens_seen": 4155968,441 "step": 270442 },443 {444 "epoch": 1.1282051282051282,445 "grad_norm": 0.421875,446 "learning_rate": 2.9907515893368784e-06,447 "loss": 0.537,448 "num_input_tokens_seen": 4234272,449 "step": 275450 },451 {452 "epoch": 1.1487179487179486,453 "grad_norm": 0.40234375,454 "learning_rate": 2.9904125881362378e-06,455 "loss": 0.5305,456 "num_input_tokens_seen": 4302368,457 "step": 280458 },459 {460 "epoch": 1.1692307692307693,461 "grad_norm": 0.5625,462 "learning_rate": 2.990067505114896e-06,463 "loss": 0.5337,464 "num_input_tokens_seen": 4376640,465 "step": 285466 },467 {468 "epoch": 1.1897435897435897,469 "grad_norm": 0.451171875,470 "learning_rate": 2.9897163416810084e-06,471 "loss": 0.592,472 "num_input_tokens_seen": 4458208,473 "step": 290474 },475 {476 "epoch": 1.2102564102564102,477 "grad_norm": 0.40625,478 "learning_rate": 2.9893590992675427e-06,479 "loss": 0.5808,480 "num_input_tokens_seen": 4537920,481 "step": 295482 },483 {484 "epoch": 1.2307692307692308,485 "grad_norm": 0.58984375,486 "learning_rate": 2.988995779332273e-06,487 "loss": 0.5569,488 "num_input_tokens_seen": 4606880,489 "step": 300490 },491 {492 "epoch": 1.2512820512820513,493 "grad_norm": 0.431640625,494 "learning_rate": 2.9886263833577725e-06,495 "loss": 0.5422,496 "num_input_tokens_seen": 4682816,497 "step": 305498 },499 {500 "epoch": 1.2717948717948717,501 "grad_norm": 0.5234375,502 "learning_rate": 2.98825091285141e-06,503 "loss": 0.5296,504 "num_input_tokens_seen": 4756032,505 "step": 310506 },507 {508 "epoch": 1.2923076923076924,509 "grad_norm": 0.4453125,510 "learning_rate": 2.987869369345341e-06,511 "loss": 0.5012,512 "num_input_tokens_seen": 4827232,513 "step": 315514 },515 {516 "epoch": 1.3128205128205128,517 "grad_norm": 0.392578125,518 "learning_rate": 2.987481754396502e-06,519 "loss": 0.5289,520 "num_input_tokens_seen": 4902368,521 "step": 320522 },523 {524 "epoch": 1.3333333333333333,525 "grad_norm": 0.326171875,526 "learning_rate": 2.9870880695866067e-06,527 "loss": 0.5245,528 "num_input_tokens_seen": 4978080,529 "step": 325530 },531 {532 "epoch": 1.353846153846154,533 "grad_norm": 0.62109375,534 "learning_rate": 2.986688316522136e-06,535 "loss": 0.5325,536 "num_input_tokens_seen": 5047456,537 "step": 330538 },539 {540 "epoch": 1.3743589743589744,541 "grad_norm": 0.390625,542 "learning_rate": 2.9862824968343352e-06,543 "loss": 0.5068,544 "num_input_tokens_seen": 5118720,545 "step": 335546 },547 {548 "epoch": 1.3948717948717948,549 "grad_norm": 0.345703125,550 "learning_rate": 2.9858706121792036e-06,551 "loss": 0.5165,552 "num_input_tokens_seen": 5196288,553 "step": 340554 },555 {556 "epoch": 1.4153846153846155,557 "grad_norm": 0.251953125,558 "learning_rate": 2.985452664237488e-06,559 "loss": 0.5025,560 "num_input_tokens_seen": 5272480,561 "step": 345562 },563 {564 "epoch": 1.435897435897436,565 "grad_norm": 0.302734375,566 "learning_rate": 2.98502865471468e-06,567 "loss": 0.5285,568 "num_input_tokens_seen": 5343296,569 "step": 350570 },571 {572 "epoch": 1.4564102564102563,573 "grad_norm": 0.314453125,574 "learning_rate": 2.9845985853410053e-06,575 "loss": 0.4983,576 "num_input_tokens_seen": 5415904,577 "step": 355578 },579 {580 "epoch": 1.476923076923077,581 "grad_norm": 0.236328125,582 "learning_rate": 2.9841624578714167e-06,583 "loss": 0.5789,584 "num_input_tokens_seen": 5502176,585 "step": 360586 },587 {588 "epoch": 1.4974358974358974,589 "grad_norm": 0.2060546875,590 "learning_rate": 2.9837202740855897e-06,591 "loss": 0.5394,592 "num_input_tokens_seen": 5580352,593 "step": 365594 },595 {596 "epoch": 1.5179487179487179,597 "grad_norm": 0.267578125,598 "learning_rate": 2.9832720357879107e-06,599 "loss": 0.5664,600 "num_input_tokens_seen": 5662912,601 "step": 370602 },603 {604 "epoch": 1.5384615384615383,605 "grad_norm": 0.22265625,606 "learning_rate": 2.9828177448074753e-06,607 "loss": 0.5546,608 "num_input_tokens_seen": 5743776,609 "step": 375610 },611 {612 "epoch": 1.558974358974359,613 "grad_norm": 0.20703125,614 "learning_rate": 2.9823574029980757e-06,615 "loss": 0.5412,616 "num_input_tokens_seen": 5812384,617 "step": 380618 },619 {620 "epoch": 1.5794871794871796,621 "grad_norm": 0.1630859375,622 "learning_rate": 2.981891012238196e-06,623 "loss": 0.5587,624 "num_input_tokens_seen": 5892768,625 "step": 385626 },627 {628 "epoch": 1.6,629 "grad_norm": 0.1640625,630 "learning_rate": 2.9814185744310036e-06,631 "loss": 0.535,632 "num_input_tokens_seen": 5974592,633 "step": 390634 },635 {636 "epoch": 1.6205128205128205,637 "grad_norm": 0.1533203125,638 "learning_rate": 2.9809400915043424e-06,639 "loss": 0.512,640 "num_input_tokens_seen": 6050080,641 "step": 395642 },643 {644 "epoch": 1.641025641025641,645 "grad_norm": 0.142578125,646 "learning_rate": 2.9804555654107243e-06,647 "loss": 0.5392,648 "num_input_tokens_seen": 6137248,649 "step": 400650 },651 {652 "epoch": 1.6615384615384614,653 "grad_norm": 0.1376953125,654 "learning_rate": 2.9799649981273185e-06,655 "loss": 0.6444,656 "num_input_tokens_seen": 6232704,657 "step": 405658 },659 {660 "epoch": 1.682051282051282,661 "grad_norm": 0.1435546875,662 "learning_rate": 2.9794683916559493e-06,663 "loss": 0.5202,664 "num_input_tokens_seen": 6304064,665 "step": 410666 },667 {668 "epoch": 1.7025641025641025,669 "grad_norm": 0.1455078125,670 "learning_rate": 2.9789657480230842e-06,671 "loss": 0.5344,672 "num_input_tokens_seen": 6392928,673 "step": 415674 },675 {676 "epoch": 1.7230769230769232,677 "grad_norm": 0.16015625,678 "learning_rate": 2.9784570692798236e-06,679 "loss": 0.4614,680 "num_input_tokens_seen": 6473120,681 "step": 420682 },683 {684 "epoch": 1.7435897435897436,685 "grad_norm": 0.19921875,686 "learning_rate": 2.977942357501898e-06,687 "loss": 0.5036,688 "num_input_tokens_seen": 6545312,689 "step": 425690 },691 {692 "epoch": 1.764102564102564,693 "grad_norm": 0.41796875,694 "learning_rate": 2.977421614789655e-06,695 "loss": 0.5308,696 "num_input_tokens_seen": 6629984,697 "step": 430698 },699 {700 "epoch": 1.7846153846153845,701 "grad_norm": 0.1767578125,702 "learning_rate": 2.976894843268051e-06,703 "loss": 0.5475,704 "num_input_tokens_seen": 6715936,705 "step": 435706 },707 {708 "epoch": 1.8051282051282052,709 "grad_norm": 0.150390625,710 "learning_rate": 2.976362045086647e-06,711 "loss": 0.5704,712 "num_input_tokens_seen": 6797472,713 "step": 440714 },715 {716 "epoch": 1.8256410256410256,717 "grad_norm": 0.416015625,718 "learning_rate": 2.975823222419594e-06,719 "loss": 0.4781,720 "num_input_tokens_seen": 6867808,721 "step": 445722 },723 {724 "epoch": 1.8461538461538463,725 "grad_norm": 0.25,726 "learning_rate": 2.9752783774656267e-06,727 "loss": 0.5247,728 "num_input_tokens_seen": 6944480,729 "step": 450730 },731 {732 "epoch": 1.8666666666666667,733 "grad_norm": 0.703125,734 "learning_rate": 2.974727512448056e-06,735 "loss": 0.5161,736 "num_input_tokens_seen": 7024064,737 "step": 455738 },739 {740 "epoch": 1.8871794871794871,741 "grad_norm": 0.65234375,742 "learning_rate": 2.974170629614757e-06,743 "loss": 0.5049,744 "num_input_tokens_seen": 7103488,745 "step": 460746 },747 {748 "epoch": 1.9076923076923076,749 "grad_norm": 0.734375,750 "learning_rate": 2.9736077312381624e-06,751 "loss": 0.5712,752 "num_input_tokens_seen": 7190304,753 "step": 465754 },755 {756 "epoch": 1.9282051282051282,757 "grad_norm": 0.40234375,758 "learning_rate": 2.9730388196152513e-06,759 "loss": 0.5222,760 "num_input_tokens_seen": 7265056,761 "step": 470762 },763 {764 "epoch": 1.9487179487179487,765 "grad_norm": 0.28125,766 "learning_rate": 2.972463897067541e-06,767 "loss": 0.4829,768 "num_input_tokens_seen": 7336992,769 "step": 475770 },771 {772 "epoch": 1.9692307692307693,773 "grad_norm": 0.56640625,774 "learning_rate": 2.971882965941077e-06,775 "loss": 0.5218,776 "num_input_tokens_seen": 7413984,777 "step": 480778 },779 {780 "epoch": 1.9897435897435898,781 "grad_norm": 0.357421875,782 "learning_rate": 2.9712960286064237e-06,783 "loss": 0.5543,784 "num_input_tokens_seen": 7492768,785 "step": 485786 },787 {788 "epoch": 2.01025641025641,789 "grad_norm": 0.26171875,790 "learning_rate": 2.970703087458655e-06,791 "loss": 0.5036,792 "num_input_tokens_seen": 7563904,793 "step": 490794 },795 {796 "epoch": 2.0307692307692307,797 "grad_norm": 0.38671875,798 "learning_rate": 2.9701041449173426e-06,799 "loss": 0.556,800 "num_input_tokens_seen": 7634464,801 "step": 495802 },803 {804 "epoch": 2.051282051282051,805 "grad_norm": 0.375,806 "learning_rate": 2.96949920342655e-06,807 "loss": 0.5742,808 "num_input_tokens_seen": 7711168,809 "step": 500810 },811 {812 "epoch": 2.071794871794872,813 "grad_norm": 0.365234375,814 "learning_rate": 2.968888265454818e-06,815 "loss": 0.4905,816 "num_input_tokens_seen": 7786656,817 "step": 505818 },819 {820 "epoch": 2.0923076923076924,821 "grad_norm": 0.271484375,822 "learning_rate": 2.968271333495158e-06,823 "loss": 0.5134,824 "num_input_tokens_seen": 7858240,825 "step": 510826 },827 {828 "epoch": 2.112820512820513,829 "grad_norm": 0.373046875,830 "learning_rate": 2.967648410065041e-06,831 "loss": 0.6211,832 "num_input_tokens_seen": 7952064,833 "step": 515834 },835 {836 "epoch": 2.1333333333333333,837 "grad_norm": 0.45703125,838 "learning_rate": 2.9670194977063857e-06,839 "loss": 0.515,840 "num_input_tokens_seen": 8026688,841 "step": 520842 },843 {844 "epoch": 2.1538461538461537,845 "grad_norm": 0.400390625,846 "learning_rate": 2.96638459898555e-06,847 "loss": 0.5464,848 "num_input_tokens_seen": 8104064,849 "step": 525850 },851 {852 "epoch": 2.174358974358974,853 "grad_norm": 0.421875,854 "learning_rate": 2.9657437164933205e-06,855 "loss": 0.5331,856 "num_input_tokens_seen": 8187968,857 "step": 530858 },859 {860 "epoch": 2.194871794871795,861 "grad_norm": 0.396484375,862 "learning_rate": 2.9650968528449e-06,863 "loss": 0.5236,864 "num_input_tokens_seen": 8261312,865 "step": 535866 },867 {868 "epoch": 2.2153846153846155,869 "grad_norm": 0.34765625,870 "learning_rate": 2.9644440106799e-06,871 "loss": 0.4345,872 "num_input_tokens_seen": 8336032,873 "step": 540874 },875 {876 "epoch": 2.235897435897436,877 "grad_norm": 0.353515625,878 "learning_rate": 2.963785192662327e-06,879 "loss": 0.4853,880 "num_input_tokens_seen": 8405024,881 "step": 545882 },883 {884 "epoch": 2.2564102564102564,885 "grad_norm": 0.365234375,886 "learning_rate": 2.9631204014805716e-06,887 "loss": 0.5,888 "num_input_tokens_seen": 8483456,889 "step": 550890 },891 {892 "epoch": 2.276923076923077,893 "grad_norm": 0.38671875,894 "learning_rate": 2.9624496398474014e-06,895 "loss": 0.4863,896 "num_input_tokens_seen": 8560768,897 "step": 555898 },899 {900 "epoch": 2.2974358974358973,901 "grad_norm": 0.32421875,902 "learning_rate": 2.961772910499945e-06,903 "loss": 0.48,904 "num_input_tokens_seen": 8637888,905 "step": 560906 },907 {908 "epoch": 2.3179487179487177,909 "grad_norm": 0.376953125,910 "learning_rate": 2.9610902161996838e-06,911 "loss": 0.5768,912 "num_input_tokens_seen": 8715360,913 "step": 565914 },915 {916 "epoch": 2.3384615384615386,917 "grad_norm": 0.265625,918 "learning_rate": 2.96040155973244e-06,919 "loss": 0.4751,920 "num_input_tokens_seen": 8795200,921 "step": 570922 },923 {924 "epoch": 2.358974358974359,925 "grad_norm": 0.283203125,926 "learning_rate": 2.959706943908365e-06,927 "loss": 0.5161,928 "num_input_tokens_seen": 8883136,929 "step": 575930 },931 {932 "epoch": 2.3794871794871795,933 "grad_norm": 0.271484375,934 "learning_rate": 2.9590063715619287e-06,935 "loss": 0.5588,936 "num_input_tokens_seen": 8961568,937 "step": 580938 },939 {940 "epoch": 2.4,941 "grad_norm": 0.2373046875,942 "learning_rate": 2.9582998455519062e-06,943 "loss": 0.5527,944 "num_input_tokens_seen": 9043360,945 "step": 585946 },947 {948 "epoch": 2.4205128205128204,949 "grad_norm": 0.21875,950 "learning_rate": 2.9575873687613676e-06,951 "loss": 0.4897,952 "num_input_tokens_seen": 9116448,953 "step": 590954 },955 {956 "epoch": 2.4410256410256412,957 "grad_norm": 0.21484375,958 "learning_rate": 2.9568689440976676e-06,959 "loss": 0.5359,960 "num_input_tokens_seen": 9193120,961 "step": 595962 },963 {964 "epoch": 2.4615384615384617,965 "grad_norm": 0.205078125,966 "learning_rate": 2.95614457449243e-06,967 "loss": 0.5763,968 "num_input_tokens_seen": 9269920,969 "step": 600970 },971 {972 "epoch": 2.482051282051282,973 "grad_norm": 0.1845703125,974 "learning_rate": 2.9554142629015382e-06,975 "loss": 0.4631,976 "num_input_tokens_seen": 9339968,977 "step": 605978 },979 {980 "epoch": 2.5025641025641026,981 "grad_norm": 0.1650390625,982 "learning_rate": 2.954678012305123e-06,983 "loss": 0.5349,984 "num_input_tokens_seen": 9426976,985 "step": 610986 },987 {988 "epoch": 2.523076923076923,989 "grad_norm": 0.1708984375,990 "learning_rate": 2.9539358257075495e-06,991 "loss": 0.5532,992 "num_input_tokens_seen": 9509056,993 "step": 615994 },995 {996 "epoch": 2.5435897435897434,997 "grad_norm": 0.1611328125,998 "learning_rate": 2.9531877061374066e-06,999 "loss": 0.4748,1000 "num_input_tokens_seen": 9590720,1001 "step": 6201002 },1003 {1004 "epoch": 2.564102564102564,1005 "grad_norm": 0.15234375,1006 "learning_rate": 2.9524336566474915e-06,1007 "loss": 0.5022,1008 "num_input_tokens_seen": 9667648,1009 "step": 6251010 },1011 {1012 "epoch": 2.5846153846153848,1013 "grad_norm": 0.1337890625,1014 "learning_rate": 2.9516736803148014e-06,1015 "loss": 0.5005,1016 "num_input_tokens_seen": 9738016,1017 "step": 6301018 },1019 {1020 "epoch": 2.605128205128205,1021 "grad_norm": 0.1337890625,1022 "learning_rate": 2.9509077802405174e-06,1023 "loss": 0.5297,1024 "num_input_tokens_seen": 9816224,1025 "step": 6351026 },1027 {1028 "epoch": 2.6256410256410256,1029 "grad_norm": 0.13671875,1030 "learning_rate": 2.9501359595499933e-06,1031 "loss": 0.5399,1032 "num_input_tokens_seen": 9891104,1033 "step": 6401034 },1035 {1036 "epoch": 2.646153846153846,1037 "grad_norm": 0.177734375,1038 "learning_rate": 2.9493582213927425e-06,1039 "loss": 0.4901,1040 "num_input_tokens_seen": 9969792,1041 "step": 6451042 },1043 {1044 "epoch": 2.6666666666666665,1045 "grad_norm": 0.142578125,1046 "learning_rate": 2.9485745689424267e-06,1047 "loss": 0.4591,1048 "num_input_tokens_seen": 10044608,1049 "step": 6501050 },1051 {1052 "epoch": 2.6871794871794874,1053 "grad_norm": 0.2451171875,1054 "learning_rate": 2.9477850053968405e-06,1055 "loss": 0.5729,1056 "num_input_tokens_seen": 10132640,1057 "step": 6551058 },1059 {1060 "epoch": 2.707692307692308,1061 "grad_norm": 0.150390625,1062 "learning_rate": 2.9469895339778995e-06,1063 "loss": 0.5405,1064 "num_input_tokens_seen": 10207968,1065 "step": 6601066 },1067 {1068 "epoch": 2.7282051282051283,1069 "grad_norm": 0.30859375,1070 "learning_rate": 2.946188157931627e-06,1071 "loss": 0.4786,1072 "num_input_tokens_seen": 10277408,1073 "step": 6651074 },1075 {1076 "epoch": 2.7487179487179487,1077 "grad_norm": 0.314453125,1078 "learning_rate": 2.9453808805281423e-06,1079 "loss": 0.5035,1080 "num_input_tokens_seen": 10349184,1081 "step": 6701082 },1083 {1084 "epoch": 2.769230769230769,1085 "grad_norm": 0.2490234375,1086 "learning_rate": 2.944567705061644e-06,1087 "loss": 0.4719,1088 "num_input_tokens_seen": 10434112,1089 "step": 6751090 },1091 {1092 "epoch": 2.7897435897435896,1093 "grad_norm": 0.65234375,1094 "learning_rate": 2.9437486348504e-06,1095 "loss": 0.5118,1096 "num_input_tokens_seen": 10506208,1097 "step": 6801098 },1099 {1100 "epoch": 2.81025641025641,1101 "grad_norm": 0.68359375,1102 "learning_rate": 2.9429236732367318e-06,1103 "loss": 0.5014,1104 "num_input_tokens_seen": 10577696,1105 "step": 6851106 },1107 {1108 "epoch": 2.830769230769231,1109 "grad_norm": 0.71484375,1110 "learning_rate": 2.942092823587001e-06,1111 "loss": 0.4827,1112 "num_input_tokens_seen": 10657984,1113 "step": 6901114 },1115 {1116 "epoch": 2.8512820512820514,1117 "grad_norm": 0.71484375,1118 "learning_rate": 2.941256089291597e-06,1119 "loss": 0.5177,1120 "num_input_tokens_seen": 10734688,1121 "step": 6951122 },1123 {1124 "epoch": 2.871794871794872,1125 "grad_norm": 0.6328125,1126 "learning_rate": 2.940413473764923e-06,1127 "loss": 0.4517,1128 "num_input_tokens_seen": 10812640,1129 "step": 7001130 },1131 {1132 "epoch": 2.8923076923076922,1133 "grad_norm": 0.70703125,1134 "learning_rate": 2.9395649804453786e-06,1135 "loss": 0.4574,1136 "num_input_tokens_seen": 10884800,1137 "step": 7051138 },1139 {1140 "epoch": 2.9128205128205127,1141 "grad_norm": 0.27734375,1142 "learning_rate": 2.9387106127953515e-06,1143 "loss": 0.5092,1144 "num_input_tokens_seen": 10962016,1145 "step": 7101146 },1147 {1148 "epoch": 2.9333333333333336,1149 "grad_norm": 0.34375,1150 "learning_rate": 2.937850374301198e-06,1151 "loss": 0.4888,1152 "num_input_tokens_seen": 11033280,1153 "step": 7151154 },1155 {1156 "epoch": 2.953846153846154,1157 "grad_norm": 0.375,1158 "learning_rate": 2.9369842684732336e-06,1159 "loss": 0.5447,1160 "num_input_tokens_seen": 11113696,1161 "step": 7201162 },1163 {1164 "epoch": 2.9743589743589745,1165 "grad_norm": 0.396484375,1166 "learning_rate": 2.936112298845713e-06,1167 "loss": 0.5438,1168 "num_input_tokens_seen": 11195104,1169 "step": 7251170 },1171 {1172 "epoch": 2.994871794871795,1173 "grad_norm": 0.2470703125,1174 "learning_rate": 2.935234468976822e-06,1175 "loss": 0.46,1176 "num_input_tokens_seen": 11270304,1177 "step": 7301178 },1179 {1180 "epoch": 3.0153846153846153,1181 "grad_norm": 0.392578125,1182 "learning_rate": 2.934350782448658e-06,1183 "loss": 0.569,1184 "num_input_tokens_seen": 11350784,1185 "step": 7351186 },1187 {1188 "epoch": 3.0358974358974358,1189 "grad_norm": 0.318359375,1190 "learning_rate": 2.9334612428672175e-06,1191 "loss": 0.5246,1192 "num_input_tokens_seen": 11429568,1193 "step": 7401194 },1195 {1196 "epoch": 3.056410256410256,1197 "grad_norm": 0.35546875,1198 "learning_rate": 2.9325658538623822e-06,1199 "loss": 0.4587,1200 "num_input_tokens_seen": 11502784,