charent/self_cognition_Alice
16.4k
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 2.0,6 "eval_steps": 500,7 "global_step": 230,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.008695652173913044,14 "grad_norm": 17.666513442993164,15 "learning_rate": 0.0,16 "loss": 4.5266,17 "step": 118 },19 {20 "epoch": 0.017391304347826087,21 "grad_norm": 21.562305450439453,22 "learning_rate": 1e-05,23 "loss": 4.6683,24 "step": 225 },26 {27 "epoch": 0.02608695652173913,28 "grad_norm": 17.93792152404785,29 "learning_rate": 2e-05,30 "loss": 4.4263,31 "step": 332 },33 {34 "epoch": 0.034782608695652174,35 "grad_norm": 31.087865829467773,36 "learning_rate": 3e-05,37 "loss": 5.3455,38 "step": 439 },40 {41 "epoch": 0.043478260869565216,42 "grad_norm": 16.84352684020996,43 "learning_rate": 4e-05,44 "loss": 4.7065,45 "step": 546 },47 {48 "epoch": 0.05217391304347826,49 "grad_norm": 15.276086807250977,50 "learning_rate": 5e-05,51 "loss": 3.5936,52 "step": 653 },54 {55 "epoch": 0.06086956521739131,56 "grad_norm": 15.930822372436523,57 "learning_rate": 6e-05,58 "loss": 3.7262,59 "step": 760 },61 {62 "epoch": 0.06956521739130435,63 "grad_norm": 16.219757080078125,64 "learning_rate": 7e-05,65 "loss": 4.0023,66 "step": 867 },68 {69 "epoch": 0.0782608695652174,70 "grad_norm": 13.747424125671387,71 "learning_rate": 8e-05,72 "loss": 3.8195,73 "step": 974 },75 {76 "epoch": 0.08695652173913043,77 "grad_norm": 28.324934005737305,78 "learning_rate": 9e-05,79 "loss": 4.1877,80 "step": 1081 },82 {83 "epoch": 0.09565217391304348,84 "grad_norm": 11.464014053344727,85 "learning_rate": 0.0001,86 "loss": 2.5265,87 "step": 1188 },89 {90 "epoch": 0.10434782608695652,91 "grad_norm": 11.093965530395508,92 "learning_rate": 9.999490215047167e-05,93 "loss": 2.0965,94 "step": 1295 },96 {97 "epoch": 0.11304347826086956,98 "grad_norm": 13.068178176879883,99 "learning_rate": 9.997960964140947e-05,100 "loss": 2.4444,101 "step": 13102 },103 {104 "epoch": 0.12173913043478261,105 "grad_norm": 12.251531600952148,106 "learning_rate": 9.995412559116979e-05,107 "loss": 2.6647,108 "step": 14109 },110 {111 "epoch": 0.13043478260869565,112 "grad_norm": 15.61241340637207,113 "learning_rate": 9.991845519630678e-05,114 "loss": 2.9974,115 "step": 15116 },117 {118 "epoch": 0.1391304347826087,119 "grad_norm": 11.974143981933594,120 "learning_rate": 9.987260573051269e-05,121 "loss": 1.9998,122 "step": 16123 },124 {125 "epoch": 0.14782608695652175,126 "grad_norm": 11.015225410461426,127 "learning_rate": 9.981658654313457e-05,128 "loss": 2.1787,129 "step": 17130 },131 {132 "epoch": 0.1565217391304348,133 "grad_norm": 10.231270790100098,134 "learning_rate": 9.975040905726798e-05,135 "loss": 1.3827,136 "step": 18137 },138 {139 "epoch": 0.16521739130434782,140 "grad_norm": 10.304975509643555,141 "learning_rate": 9.967408676742751e-05,142 "loss": 1.8596,143 "step": 19144 },145 {146 "epoch": 0.17391304347826086,147 "grad_norm": 11.838497161865234,148 "learning_rate": 9.958763523679514e-05,149 "loss": 1.8492,150 "step": 20151 },152 {153 "epoch": 0.1826086956521739,154 "grad_norm": 14.471789360046387,155 "learning_rate": 9.949107209404665e-05,156 "loss": 1.4539,157 "step": 21158 },159 {160 "epoch": 0.19130434782608696,161 "grad_norm": 13.361359596252441,162 "learning_rate": 9.938441702975689e-05,163 "loss": 2.3912,164 "step": 22165 },166 {167 "epoch": 0.2,168 "grad_norm": 9.12480640411377,169 "learning_rate": 9.926769179238466e-05,170 "loss": 1.188,171 "step": 23172 },173 {174 "epoch": 0.20869565217391303,175 "grad_norm": 11.180279731750488,176 "learning_rate": 9.914092018383778e-05,177 "loss": 1.6116,178 "step": 24179 },180 {181 "epoch": 0.21739130434782608,182 "grad_norm": 11.030242919921875,183 "learning_rate": 9.900412805461967e-05,184 "loss": 1.9637,185 "step": 25186 },187 {188 "epoch": 0.22608695652173913,189 "grad_norm": 9.22864818572998,190 "learning_rate": 9.885734329855798e-05,191 "loss": 1.8468,192 "step": 26193 },194 {195 "epoch": 0.23478260869565218,196 "grad_norm": 14.804508209228516,197 "learning_rate": 9.870059584711668e-05,198 "loss": 0.9744,199 "step": 27200 },201 {202 "epoch": 0.24347826086956523,203 "grad_norm": 11.706965446472168,204 "learning_rate": 9.853391766329263e-05,205 "loss": 1.9317,206 "step": 28207 },208 {209 "epoch": 0.25217391304347825,210 "grad_norm": 11.159095764160156,211 "learning_rate": 9.835734273509786e-05,212 "loss": 1.7246,213 "step": 29214 },215 {216 "epoch": 0.2608695652173913,217 "grad_norm": 8.815181732177734,218 "learning_rate": 9.817090706862895e-05,219 "loss": 1.2355,220 "step": 30221 },222 {223 "epoch": 0.26956521739130435,224 "grad_norm": 11.841676712036133,225 "learning_rate": 9.797464868072488e-05,226 "loss": 3.3564,227 "step": 31228 },229 {230 "epoch": 0.2782608695652174,231 "grad_norm": 15.580975532531738,232 "learning_rate": 9.776860759121484e-05,233 "loss": 2.0827,234 "step": 32235 },236 {237 "epoch": 0.28695652173913044,238 "grad_norm": 8.459304809570312,239 "learning_rate": 9.755282581475769e-05,240 "loss": 1.2298,241 "step": 33242 },243 {244 "epoch": 0.2956521739130435,245 "grad_norm": 9.495912551879883,246 "learning_rate": 9.73273473522745e-05,247 "loss": 1.3418,248 "step": 34249 },250 {251 "epoch": 0.30434782608695654,252 "grad_norm": 13.812179565429688,253 "learning_rate": 9.709221818197624e-05,254 "loss": 1.6258,255 "step": 35256 },257 {258 "epoch": 0.3130434782608696,259 "grad_norm": 9.413185119628906,260 "learning_rate": 9.68474862499881e-05,261 "loss": 1.168,262 "step": 36263 },264 {265 "epoch": 0.3217391304347826,266 "grad_norm": 13.937978744506836,267 "learning_rate": 9.659320146057262e-05,268 "loss": 1.4273,269 "step": 37270 },271 {272 "epoch": 0.33043478260869563,273 "grad_norm": 9.947927474975586,274 "learning_rate": 9.632941566595357e-05,275 "loss": 1.3857,276 "step": 38277 },278 {279 "epoch": 0.3391304347826087,280 "grad_norm": 10.392467498779297,281 "learning_rate": 9.60561826557425e-05,282 "loss": 0.7599,283 "step": 39284 },285 {286 "epoch": 0.34782608695652173,287 "grad_norm": 12.23806381225586,288 "learning_rate": 9.577355814597031e-05,289 "loss": 2.0868,290 "step": 40291 },292 {293 "epoch": 0.3565217391304348,294 "grad_norm": 6.938766002655029,295 "learning_rate": 9.548159976772592e-05,296 "loss": 0.4847,297 "step": 41298 },299 {300 "epoch": 0.3652173913043478,301 "grad_norm": 9.958048820495605,302 "learning_rate": 9.518036705540458e-05,303 "loss": 1.609,304 "step": 42305 },306 {307 "epoch": 0.3739130434782609,308 "grad_norm": 11.910273551940918,309 "learning_rate": 9.486992143456792e-05,310 "loss": 0.7905,311 "step": 43312 },313 {314 "epoch": 0.3826086956521739,315 "grad_norm": 8.610260963439941,316 "learning_rate": 9.45503262094184e-05,317 "loss": 1.4258,318 "step": 44319 },320 {321 "epoch": 0.391304347826087,322 "grad_norm": 7.331972122192383,323 "learning_rate": 9.422164654989072e-05,324 "loss": 0.9974,325 "step": 45326 },327 {328 "epoch": 0.4,329 "grad_norm": 8.997504234313965,330 "learning_rate": 9.388394947836279e-05,331 "loss": 1.1141,332 "step": 46333 },334 {335 "epoch": 0.40869565217391307,336 "grad_norm": 8.395123481750488,337 "learning_rate": 9.353730385598887e-05,338 "loss": 0.9686,339 "step": 47340 },341 {342 "epoch": 0.41739130434782606,343 "grad_norm": 8.04129695892334,344 "learning_rate": 9.318178036865785e-05,345 "loss": 0.3619,346 "step": 48347 },348 {349 "epoch": 0.4260869565217391,350 "grad_norm": 5.985868453979492,351 "learning_rate": 9.281745151257946e-05,352 "loss": 0.6158,353 "step": 49354 },355 {356 "epoch": 0.43478260869565216,357 "grad_norm": 12.913836479187012,358 "learning_rate": 9.244439157950114e-05,359 "loss": 1.6131,360 "step": 50361 },362 {363 "epoch": 0.4434782608695652,364 "grad_norm": 7.429251194000244,365 "learning_rate": 9.206267664155907e-05,366 "loss": 0.8433,367 "step": 51368 },369 {370 "epoch": 0.45217391304347826,371 "grad_norm": 8.468223571777344,372 "learning_rate": 9.167238453576589e-05,373 "loss": 0.8761,374 "step": 52375 },376 {377 "epoch": 0.4608695652173913,378 "grad_norm": 10.999699592590332,379 "learning_rate": 9.12735948481387e-05,380 "loss": 1.7577,381 "step": 53382 },383 {384 "epoch": 0.46956521739130436,385 "grad_norm": 8.069478988647461,386 "learning_rate": 9.086638889747035e-05,387 "loss": 1.1008,388 "step": 54389 },390 {391 "epoch": 0.4782608695652174,392 "grad_norm": 9.439266204833984,393 "learning_rate": 9.045084971874738e-05,394 "loss": 0.7977,395 "step": 55396 },397 {398 "epoch": 0.48695652173913045,399 "grad_norm": 8.834954261779785,400 "learning_rate": 9.002706204621803e-05,401 "loss": 1.2913,402 "step": 56403 },404 {405 "epoch": 0.4956521739130435,406 "grad_norm": 14.728219032287598,407 "learning_rate": 8.959511229611376e-05,408 "loss": 1.566,409 "step": 57410 },411 {412 "epoch": 0.5043478260869565,413 "grad_norm": 14.64725112915039,414 "learning_rate": 8.915508854902778e-05,415 "loss": 0.7095,416 "step": 58417 },418 {419 "epoch": 0.5130434782608696,420 "grad_norm": 17.302316665649414,421 "learning_rate": 8.870708053195413e-05,422 "loss": 1.1349,423 "step": 59424 },425 {426 "epoch": 0.5217391304347826,427 "grad_norm": 13.19262981414795,428 "learning_rate": 8.825117959999116e-05,429 "loss": 1.0779,430 "step": 60431 },432 {433 "epoch": 0.5304347826086957,434 "grad_norm": 10.02542495727539,435 "learning_rate": 8.778747871771292e-05,436 "loss": 0.7204,437 "step": 61438 },439 {440 "epoch": 0.5391304347826087,441 "grad_norm": 2.038048505783081,442 "learning_rate": 8.731607244021236e-05,443 "loss": 0.0617,444 "step": 62445 },446 {447 "epoch": 0.5478260869565217,448 "grad_norm": 8.415672302246094,449 "learning_rate": 8.683705689382024e-05,450 "loss": 1.199,451 "step": 63452 },453 {454 "epoch": 0.5565217391304348,455 "grad_norm": 10.012858390808105,456 "learning_rate": 8.635052975650369e-05,457 "loss": 1.442,458 "step": 64459 },460 {461 "epoch": 0.5652173913043478,462 "grad_norm": 10.651287078857422,463 "learning_rate": 8.585659023794818e-05,464 "loss": 1.6201,465 "step": 65466 },467 {468 "epoch": 0.5739130434782609,469 "grad_norm": 8.589141845703125,470 "learning_rate": 8.535533905932738e-05,471 "loss": 1.1879,472 "step": 66473 },474 {475 "epoch": 0.5826086956521739,476 "grad_norm": 9.711124420166016,477 "learning_rate": 8.484687843276469e-05,478 "loss": 1.2133,479 "step": 67480 },481 {482 "epoch": 0.591304347826087,483 "grad_norm": 9.756671905517578,484 "learning_rate": 8.433131204049067e-05,485 "loss": 1.1111,486 "step": 68487 },488 {489 "epoch": 0.6,490 "grad_norm": 10.514204025268555,491 "learning_rate": 8.380874501370097e-05,492 "loss": 0.4235,493 "step": 69494 },495 {496 "epoch": 0.6086956521739131,497 "grad_norm": 6.404147148132324,498 "learning_rate": 8.327928391111841e-05,499 "loss": 0.1837,500 "step": 70501 },502 {503 "epoch": 0.6173913043478261,504 "grad_norm": 12.888923645019531,505 "learning_rate": 8.274303669726426e-05,506 "loss": 1.2996,507 "step": 71508 },509 {510 "epoch": 0.6260869565217392,511 "grad_norm": 12.083588600158691,512 "learning_rate": 8.220011272044277e-05,513 "loss": 0.3065,514 "step": 72515 },516 {517 "epoch": 0.6347826086956522,518 "grad_norm": 7.7776665687561035,519 "learning_rate": 8.165062269044353e-05,520 "loss": 0.7543,521 "step": 73522 },523 {524 "epoch": 0.6434782608695652,525 "grad_norm": 8.809155464172363,526 "learning_rate": 8.109467865596612e-05,527 "loss": 1.0967,528 "step": 74529 },530 {531 "epoch": 0.6521739130434783,532 "grad_norm": 18.32766342163086,533 "learning_rate": 8.053239398177191e-05,534 "loss": 2.4623,535 "step": 75536 },537 {538 "epoch": 0.6608695652173913,539 "grad_norm": 12.604337692260742,540 "learning_rate": 7.996388332556735e-05,541 "loss": 1.7208,542 "step": 76543 },544 {545 "epoch": 0.6695652173913044,546 "grad_norm": 12.69120979309082,547 "learning_rate": 7.938926261462366e-05,548 "loss": 0.883,549 "step": 77550 },551 {552 "epoch": 0.6782608695652174,553 "grad_norm": 12.506765365600586,554 "learning_rate": 7.880864902213765e-05,555 "loss": 1.83,556 "step": 78557 },558 {559 "epoch": 0.6869565217391305,560 "grad_norm": 8.688498497009277,561 "learning_rate": 7.822216094333847e-05,562 "loss": 0.9735,563 "step": 79564 },565 {566 "epoch": 0.6956521739130435,567 "grad_norm": 14.368943214416504,568 "learning_rate": 7.762991797134514e-05,569 "loss": 2.193,570 "step": 80571 },572 {573 "epoch": 0.7043478260869566,574 "grad_norm": 6.735222339630127,575 "learning_rate": 7.703204087277988e-05,576 "loss": 0.6682,577 "step": 81578 },579 {580 "epoch": 0.7130434782608696,581 "grad_norm": 15.296740531921387,582 "learning_rate": 7.64286515631421e-05,583 "loss": 0.9763,584 "step": 82585 },586 {587 "epoch": 0.7217391304347827,588 "grad_norm": 16.399547576904297,589 "learning_rate": 7.58198730819481e-05,590 "loss": 2.4849,591 "step": 83592 },593 {594 "epoch": 0.7304347826086957,595 "grad_norm": 11.530505180358887,596 "learning_rate": 7.52058295676416e-05,597 "loss": 1.8405,598 "step": 84599 },600 {601 "epoch": 0.7391304347826086,602 "grad_norm": 7.331994533538818,603 "learning_rate": 7.45866462322802e-05,604 "loss": 0.9336,605 "step": 85606 },607 {608 "epoch": 0.7478260869565218,609 "grad_norm": 20.28060531616211,610 "learning_rate": 7.396244933600285e-05,611 "loss": 1.2999,612 "step": 86613 },614 {615 "epoch": 0.7565217391304347,616 "grad_norm": 8.961277961730957,617 "learning_rate": 7.333336616128369e-05,618 "loss": 1.0094,619 "step": 87620 },621 {622 "epoch": 0.7652173913043478,623 "grad_norm": 7.146012306213379,624 "learning_rate": 7.269952498697734e-05,625 "loss": 0.8734,626 "step": 88627 },628 {629 "epoch": 0.7739130434782608,630 "grad_norm": 12.372821807861328,631 "learning_rate": 7.206105506216106e-05,632 "loss": 1.5859,633 "step": 89634 },635 {636 "epoch": 0.782608695652174,637 "grad_norm": 8.59324836730957,638 "learning_rate": 7.141808657977907e-05,639 "loss": 1.0735,640 "step": 90641 },642 {643 "epoch": 0.7913043478260869,644 "grad_norm": 6.564212799072266,645 "learning_rate": 7.077075065009433e-05,646 "loss": 0.8069,647 "step": 91648 },649 {650 "epoch": 0.8,651 "grad_norm": 10.416309356689453,652 "learning_rate": 7.01191792739534e-05,653 "loss": 1.0411,654 "step": 92655 },656 {657 "epoch": 0.808695652173913,658 "grad_norm": 11.143993377685547,659 "learning_rate": 6.946350531586959e-05,660 "loss": 0.8595,661 "step": 93662 },663 {664 "epoch": 0.8173913043478261,665 "grad_norm": 9.861275672912598,666 "learning_rate": 6.880386247692999e-05,667 "loss": 0.9368,668 "step": 94669 },670 {671 "epoch": 0.8260869565217391,672 "grad_norm": 10.142037391662598,673 "learning_rate": 6.814038526753205e-05,674 "loss": 1.5361,675 "step": 95676 },677 {678 "epoch": 0.8347826086956521,679 "grad_norm": 6.418692588806152,680 "learning_rate": 6.747320897995493e-05,681 "loss": 0.7167,682 "step": 96683 },684 {685 "epoch": 0.8434782608695652,686 "grad_norm": 4.113799571990967,687 "learning_rate": 6.680246966077151e-05,688 "loss": 0.3922,689 "step": 97690 },691 {692 "epoch": 0.8521739130434782,693 "grad_norm": 10.264501571655273,694 "learning_rate": 6.61283040831067e-05,695 "loss": 2.2257,696 "step": 98697 },698 {699 "epoch": 0.8608695652173913,700 "grad_norm": 15.571903228759766,701 "learning_rate": 6.545084971874738e-05,702 "loss": 2.4521,703 "step": 99704 },705 {706 "epoch": 0.8695652173913043,707 "grad_norm": 7.295588970184326,708 "learning_rate": 6.477024471011001e-05,709 "loss": 0.7105,710 "step": 100711 },712 {713 "epoch": 0.8782608695652174,714 "grad_norm": 11.271722793579102,715 "learning_rate": 6.408662784207149e-05,716 "loss": 2.0426,717 "step": 101718 },719 {720 "epoch": 0.8869565217391304,721 "grad_norm": 8.137569427490234,722 "learning_rate": 6.340013851366896e-05,723 "loss": 1.3701,724 "step": 102725 },726 {727 "epoch": 0.8956521739130435,728 "grad_norm": 10.493950843811035,729 "learning_rate": 6.271091670967436e-05,730 "loss": 1.1864,731 "step": 103732 },733 {734 "epoch": 0.9043478260869565,735 "grad_norm": 9.262134552001953,736 "learning_rate": 6.201910297204962e-05,737 "loss": 1.0413,738 "step": 104739 },740 {741 "epoch": 0.9130434782608695,742 "grad_norm": 10.564716339111328,743 "learning_rate": 6.132483837128823e-05,744 "loss": 2.303,745 "step": 105746 },747 {748 "epoch": 0.9217391304347826,749 "grad_norm": 4.90865421295166,750 "learning_rate": 6.062826447764883e-05,751 "loss": 0.3802,752 "step": 106753 },754 {755 "epoch": 0.9304347826086956,756 "grad_norm": 9.297486305236816,757 "learning_rate": 5.992952333228728e-05,758 "loss": 1.2512,759 "step": 107760 },761 {762 "epoch": 0.9391304347826087,763 "grad_norm": 8.179728507995605,764 "learning_rate": 5.9228757418292266e-05,765 "loss": 0.4771,766 "step": 108767 },768 {769 "epoch": 0.9478260869565217,770 "grad_norm": 18.19074249267578,771 "learning_rate": 5.85261096316312e-05,772 "loss": 2.0378,773 "step": 109774 },775 {776 "epoch": 0.9565217391304348,777 "grad_norm": 4.808218955993652,778 "learning_rate": 5.782172325201155e-05,779 "loss": 0.14,780 "step": 110781 },782 {783 "epoch": 0.9652173913043478,784 "grad_norm": 15.154193878173828,785 "learning_rate": 5.7115741913664264e-05,786 "loss": 1.3785,787 "step": 111788 },789 {790 "epoch": 0.9739130434782609,791 "grad_norm": 7.520651340484619,792 "learning_rate": 5.640830957605465e-05,793 "loss": 0.7942,794 "step": 112795 },796 {797 "epoch": 0.9826086956521739,798 "grad_norm": 11.108746528625488,799 "learning_rate": 5.569957049452703e-05,800 "loss": 2.0485,801 "step": 113802 },803 {804 "epoch": 0.991304347826087,805 "grad_norm": 8.641176223754883,806 "learning_rate": 5.4989669190889136e-05,807 "loss": 1.1212,808 "step": 114809 },810 {811 "epoch": 1.0,812 "grad_norm": 6.526795387268066,813 "learning_rate": 5.427875042394199e-05,814 "loss": 0.7003,815 "step": 115816 },817 {818 "epoch": 1.008695652173913,819 "grad_norm": 7.168546676635742,820 "learning_rate": 5.3566959159961615e-05,821 "loss": 1.0922,822 "step": 116823 },824 {825 "epoch": 1.017391304347826,826 "grad_norm": 8.669252395629883,827 "learning_rate": 5.2854440543138406e-05,828 "loss": 0.8577,829 "step": 117830 },831 {832 "epoch": 1.0260869565217392,833 "grad_norm": 7.540567398071289,834 "learning_rate": 5.2141339865980134e-05,835 "loss": 0.7922,836 "step": 118837 },838 {839 "epoch": 1.0347826086956522,840 "grad_norm": 7.116650581359863,841 "learning_rate": 5.142780253968481e-05,842 "loss": 0.7596,843 "step": 119844 },845 {846 "epoch": 1.0434782608695652,847 "grad_norm": 8.778299331665039,848 "learning_rate": 5.0713974064489367e-05,849 "loss": 2.5862,850 "step": 120851 },852 {853 "epoch": 1.0521739130434782,854 "grad_norm": 8.101160049438477,855 "learning_rate": 5e-05,856 "loss": 0.6208,857 "step": 121858 },859 {860 "epoch": 1.0608695652173914,861 "grad_norm": 5.799407958984375,862 "learning_rate": 4.928602593551065e-05,863 "loss": 0.5619,864 "step": 122865 },866 {867 "epoch": 1.0695652173913044,868 "grad_norm": 5.000042915344238,869 "learning_rate": 4.85721974603152e-05,870 "loss": 0.2951,871 "step": 123872 },873 {874 "epoch": 1.0782608695652174,875 "grad_norm": 8.474133491516113,876 "learning_rate": 4.7858660134019884e-05,877 "loss": 1.2103,878 "step": 124879 },880 {881 "epoch": 1.0869565217391304,882 "grad_norm": 10.887007713317871,883 "learning_rate": 4.71455594568616e-05,884 "loss": 1.7797,885 "step": 125886 },887 {888 "epoch": 1.0956521739130434,889 "grad_norm": 7.625034809112549,890 "learning_rate": 4.643304084003839e-05,891 "loss": 0.7967,892 "step": 126893 },894 {895 "epoch": 1.1043478260869566,896 "grad_norm": 8.578575134277344,897 "learning_rate": 4.5721249576058027e-05,898 "loss": 0.6249,899 "step": 127900 },901 {902 "epoch": 1.1130434782608696,903 "grad_norm": 6.662336349487305,904 "learning_rate": 4.501033080911086e-05,905 "loss": 0.637,906 "step": 128907 },908 {909 "epoch": 1.1217391304347826,910 "grad_norm": 8.214466094970703,911 "learning_rate": 4.4300429505472976e-05,912 "loss": 1.2037,913 "step": 129914 },915 {916 "epoch": 1.1304347826086956,917 "grad_norm": 9.812641143798828,918 "learning_rate": 4.359169042394536e-05,919 "loss": 1.0598,920 "step": 130921 },922 {923 "epoch": 1.1391304347826088,924 "grad_norm": 5.729506969451904,925 "learning_rate": 4.288425808633575e-05,926 "loss": 0.4854,927 "step": 131928 },929 {930 "epoch": 1.1478260869565218,931 "grad_norm": 7.675601959228516,932 "learning_rate": 4.2178276747988446e-05,933 "loss": 1.6833,934 "step": 132935 },936 {937 "epoch": 1.1565217391304348,938 "grad_norm": 5.756649494171143,939 "learning_rate": 4.147389036836881e-05,940 "loss": 0.586,941 "step": 133942 },943 {944 "epoch": 1.1652173913043478,945 "grad_norm": 4.710939407348633,946 "learning_rate": 4.077124258170774e-05,947 "loss": 0.1098,948 "step": 134949 },950 {951 "epoch": 1.1739130434782608,952 "grad_norm": 8.076876640319824,953 "learning_rate": 4.007047666771274e-05,954 "loss": 0.8247,955 "step": 135956 },957 {958 "epoch": 1.182608695652174,959 "grad_norm": 6.092112064361572,960 "learning_rate": 3.937173552235117e-05,961 "loss": 0.4976,962 "step": 136963 },964 {965 "epoch": 1.191304347826087,966 "grad_norm": 12.222753524780273,967 "learning_rate": 3.8675161628711776e-05,968 "loss": 0.7614,969 "step": 137970 },971 {972 "epoch": 1.2,973 "grad_norm": 9.492755889892578,974 "learning_rate": 3.798089702795038e-05,975 "loss": 0.7278,976 "step": 138977 },978 {979 "epoch": 1.208695652173913,980 "grad_norm": 8.09307861328125,981 "learning_rate": 3.728908329032567e-05,982 "loss": 1.0424,983 "step": 139984 },985 {986 "epoch": 1.2173913043478262,987 "grad_norm": 13.750467300415039,988 "learning_rate": 3.659986148633107e-05,989 "loss": 1.3154,990 "step": 140991 },992 {993 "epoch": 1.2260869565217392,994 "grad_norm": 5.85883903503418,995 "learning_rate": 3.591337215792852e-05,996 "loss": 0.151,997 "step": 141998 },999 {1000 "epoch": 1.2347826086956522,1001 "grad_norm": 7.29556131362915,1002 "learning_rate": 3.522975528989e-05,1003 "loss": 0.4567,1004 "step": 1421005 },1006 {1007 "epoch": 1.2434782608695651,1008 "grad_norm": 10.308717727661133,1009 "learning_rate": 3.4549150281252636e-05,1010 "loss": 1.3977,1011 "step": 1431012 },1013 {1014 "epoch": 1.2521739130434781,1015 "grad_norm": 9.417724609375,1016 "learning_rate": 3.3871695916893314e-05,1017 "loss": 0.948,1018 "step": 1441019 },1020 {1021 "epoch": 1.2608695652173914,1022 "grad_norm": 8.931994438171387,1023 "learning_rate": 3.3197530339228487e-05,1024 "loss": 0.9489,1025 "step": 1451026 },1027 {1028 "epoch": 1.2695652173913043,1029 "grad_norm": 14.605033874511719,1030 "learning_rate": 3.2526791020045086e-05,1031 "loss": 1.851,1032 "step": 1461033 },1034 {1035 "epoch": 1.2782608695652173,1036 "grad_norm": 7.451152324676514,1037 "learning_rate": 3.1859614732467954e-05,1038 "loss": 0.4653,1039 "step": 1471040 },1041 {1042 "epoch": 1.2869565217391306,1043 "grad_norm": 7.847959041595459,1044 "learning_rate": 3.119613752307002e-05,1045 "loss": 0.8261,1046 "step": 1481047 },1048 {1049 "epoch": 1.2956521739130435,1050 "grad_norm": 6.851174831390381,1051 "learning_rate": 3.053649468413043e-05,1052 "loss": 0.5726,1053 "step": 1491054 },1055 {1056 "epoch": 1.3043478260869565,1057 "grad_norm": 7.77811336517334,1058 "learning_rate": 2.988082072604661e-05,1059 "loss": 0.8105,1060 "step": 1501061 },1062 {1063 "epoch": 1.3130434782608695,1064 "grad_norm": 9.027288436889648,1065 "learning_rate": 2.9229249349905684e-05,1066 "loss": 1.3896,1067 "step": 1511068 },1069 {1070 "epoch": 1.3217391304347825,1071 "grad_norm": 6.307928085327148,1072 "learning_rate": 2.858191342022095e-05,1073 "loss": 0.4174,1074 "step": 1521075 },1076 {1077 "epoch": 1.3304347826086955,1078 "grad_norm": 6.78994083404541,1079 "learning_rate": 2.7938944937838923e-05,1080 "loss": 0.5739,1081 "step": 1531082 },1083 {1084 "epoch": 1.3391304347826087,1085 "grad_norm": 7.77897834777832,1086 "learning_rate": 2.7300475013022663e-05,1087 "loss": 0.6822,1088 "step": 1541089 },1090 {1091 "epoch": 1.3478260869565217,1092 "grad_norm": 10.742020606994629,1093 "learning_rate": 2.6666633838716314e-05,1094 "loss": 1.1842,1095 "step": 1551096 },1097 {1098 "epoch": 1.3565217391304347,1099 "grad_norm": 6.034131050109863,1100 "learning_rate": 2.603755066399718e-05,1101 "loss": 0.1571,1102 "step": 1561103 },1104 {1105 "epoch": 1.365217391304348,1106 "grad_norm": 8.707799911499023,1107 "learning_rate": 2.5413353767719805e-05,1108 "loss": 0.9573,1109 "step": 1571110 },1111 {1112 "epoch": 1.373913043478261,1113 "grad_norm": 5.427333354949951,1114 "learning_rate": 2.4794170432358415e-05,1115 "loss": 0.4321,1116 "step": 1581117 },1118 {1119 "epoch": 1.382608695652174,1120 "grad_norm": 18.620357513427734,1121 "learning_rate": 2.418012691805191e-05,1122 "loss": 2.3272,1123 "step": 1591124 },1125 {1126 "epoch": 1.391304347826087,1127 "grad_norm": 10.270234107971191,1128 "learning_rate": 2.3571348436857904e-05,1129 "loss": 0.7804,1130 "step": 1601131 },1132 {1133 "epoch": 1.4,1134 "grad_norm": 4.281362056732178,1135 "learning_rate": 2.296795912722014e-05,1136 "loss": 0.1796,1137 "step": 1611138 },1139 {1140 "epoch": 1.4086956521739131,1141 "grad_norm": 6.379110336303711,1142 "learning_rate": 2.2370082028654866e-05,1143 "loss": 0.5788,1144 "step": 1621145 },1146 {1147 "epoch": 1.4173913043478261,1148 "grad_norm": 1.9616364240646362,1149 "learning_rate": 2.1777839056661554e-05,1150 "loss": 0.0406,1151 "step": 1631152 },1153 {1154 "epoch": 1.4260869565217391,1155 "grad_norm": 8.34288215637207,1156 "learning_rate": 2.119135097786236e-05,1157 "loss": 0.914,1158 "step": 1641159 },1160 {1161 "epoch": 1.434782608695652,1162 "grad_norm": 2.7565298080444336,1163 "learning_rate": 2.061073738537635e-05,1164 "loss": 0.0697,1165 "step": 1651166 },1167 {1168 "epoch": 1.4434782608695653,1169 "grad_norm": 12.554067611694336,1170 "learning_rate": 2.0036116674432654e-05,1171 "loss": 1.2686,1172 "step": 1661173 },1174 {1175 "epoch": 1.4521739130434783,1176 "grad_norm": 7.41255521774292,1177 "learning_rate": 1.946760601822809e-05,1178 "loss": 0.6045,1179 "step": 1671180 },1181 {1182 "epoch": 1.4608695652173913,1183 "grad_norm": 8.661652565002441,1184 "learning_rate": 1.8905321344033898e-05,1185 "loss": 0.2404,1186 "step": 1681187 },1188 {1189 "epoch": 1.4695652173913043,1190 "grad_norm": 8.020094871520996,1191 "learning_rate": 1.8349377309556486e-05,1192 "loss": 0.627,1193 "step": 1691194 },1195 {1196 "epoch": 1.4782608695652173,1197 "grad_norm": 6.1133222579956055,1198 "learning_rate": 1.7799887279557237e-05,1199 "loss": 0.6232,1200 "step": 170