rohinm/capability-aware-new
06
1{2 "best_metric": null,3 "best_model_checkpoint": null,4 "epoch": 1.0,5 "eval_steps": 0,6 "global_step": 7474,7 "is_hyper_param_search": false,8 "is_local_process_zero": true,9 "is_world_process_zero": true,10 "log_history": [11 {12 "epoch": 0.0001337971635001338,13 "grad_norm": 153.94268798828125,14 "learning_rate": 9.998662028364998e-07,15 "loss": 0.7547,16 "step": 117 },18 {19 "epoch": 0.0002675943270002676,20 "grad_norm": 156.0087890625,21 "learning_rate": 9.997324056729997e-07,22 "loss": 0.4447,23 "step": 224 },25 {26 "epoch": 0.0004013914905004014,27 "grad_norm": 126.08484649658203,28 "learning_rate": 9.995986085094996e-07,29 "loss": 0.5753,30 "step": 331 },32 {33 "epoch": 0.0005351886540005352,34 "grad_norm": 220.88833618164062,35 "learning_rate": 9.994648113459995e-07,36 "loss": 0.9113,37 "step": 438 },39 {40 "epoch": 0.000668985817500669,41 "grad_norm": 96.56456756591797,42 "learning_rate": 9.993310141824994e-07,43 "loss": 0.6059,44 "step": 545 },46 {47 "epoch": 0.0008027829810008028,48 "grad_norm": 66.32806396484375,49 "learning_rate": 9.991972170189992e-07,50 "loss": 0.3555,51 "step": 652 },53 {54 "epoch": 0.0009365801445009366,55 "grad_norm": 206.52401733398438,56 "learning_rate": 9.990634198554991e-07,57 "loss": 0.874,58 "step": 759 },60 {61 "epoch": 0.0010703773080010704,62 "grad_norm": 87.55257415771484,63 "learning_rate": 9.989296226919988e-07,64 "loss": 0.4952,65 "step": 866 },67 {68 "epoch": 0.001204174471501204,69 "grad_norm": 57.25172805786133,70 "learning_rate": 9.987958255284987e-07,71 "loss": 0.594,72 "step": 973 },74 {75 "epoch": 0.001337971635001338,76 "grad_norm": 161.78453063964844,77 "learning_rate": 9.986620283649985e-07,78 "loss": 0.602,79 "step": 1080 },81 {82 "epoch": 0.0014717687985014718,83 "grad_norm": 26.5186710357666,84 "learning_rate": 9.985282312014984e-07,85 "loss": 0.538,86 "step": 1187 },88 {89 "epoch": 0.0016055659620016055,90 "grad_norm": 112.02456665039062,91 "learning_rate": 9.983944340379983e-07,92 "loss": 0.5882,93 "step": 1294 },95 {96 "epoch": 0.0017393631255017393,97 "grad_norm": 37.167030334472656,98 "learning_rate": 9.982606368744982e-07,99 "loss": 0.6333,100 "step": 13101 },102 {103 "epoch": 0.0018731602890018732,104 "grad_norm": 55.95832824707031,105 "learning_rate": 9.98126839710998e-07,106 "loss": 0.6163,107 "step": 14108 },109 {110 "epoch": 0.002006957452502007,111 "grad_norm": 52.30416488647461,112 "learning_rate": 9.97993042547498e-07,113 "loss": 0.5742,114 "step": 15115 },116 {117 "epoch": 0.0021407546160021407,118 "grad_norm": 43.40435028076172,119 "learning_rate": 9.978592453839978e-07,120 "loss": 0.5529,121 "step": 16122 },123 {124 "epoch": 0.0022745517795022745,125 "grad_norm": 34.061317443847656,126 "learning_rate": 9.977254482204977e-07,127 "loss": 0.5573,128 "step": 17129 },130 {131 "epoch": 0.002408348943002408,132 "grad_norm": 21.99197769165039,133 "learning_rate": 9.975916510569976e-07,134 "loss": 0.5308,135 "step": 18136 },137 {138 "epoch": 0.002542146106502542,139 "grad_norm": 99.7840805053711,140 "learning_rate": 9.974578538934974e-07,141 "loss": 0.9071,142 "step": 19143 },144 {145 "epoch": 0.002675943270002676,146 "grad_norm": 26.018207550048828,147 "learning_rate": 9.973240567299973e-07,148 "loss": 0.4285,149 "step": 20150 },151 {152 "epoch": 0.00280974043350281,153 "grad_norm": 73.99343872070312,154 "learning_rate": 9.971902595664972e-07,155 "loss": 0.6916,156 "step": 21157 },158 {159 "epoch": 0.0029435375970029436,160 "grad_norm": 62.607662200927734,161 "learning_rate": 9.97056462402997e-07,162 "loss": 0.685,163 "step": 22164 },165 {166 "epoch": 0.0030773347605030773,167 "grad_norm": 33.15293502807617,168 "learning_rate": 9.96922665239497e-07,169 "loss": 0.5709,170 "step": 23171 },172 {173 "epoch": 0.003211131924003211,174 "grad_norm": 16.63718605041504,175 "learning_rate": 9.967888680759966e-07,176 "loss": 0.4943,177 "step": 24178 },179 {180 "epoch": 0.003344929087503345,181 "grad_norm": 40.4935302734375,182 "learning_rate": 9.966550709124965e-07,183 "loss": 0.4227,184 "step": 25185 },186 {187 "epoch": 0.0034787262510034785,188 "grad_norm": 11.906813621520996,189 "learning_rate": 9.965212737489966e-07,190 "loss": 0.5615,191 "step": 26192 },193 {194 "epoch": 0.0036125234145036127,195 "grad_norm": 28.17329216003418,196 "learning_rate": 9.963874765854965e-07,197 "loss": 0.5264,198 "step": 27199 },200 {201 "epoch": 0.0037463205780037465,202 "grad_norm": 23.820280075073242,203 "learning_rate": 9.962536794219962e-07,204 "loss": 0.6427,205 "step": 28206 },207 {208 "epoch": 0.00388011774150388,209 "grad_norm": 31.91474723815918,210 "learning_rate": 9.96119882258496e-07,211 "loss": 0.472,212 "step": 29213 },214 {215 "epoch": 0.004013914905004014,216 "grad_norm": 11.115065574645996,217 "learning_rate": 9.95986085094996e-07,218 "loss": 0.5671,219 "step": 30220 },221 {222 "epoch": 0.004147712068504148,223 "grad_norm": 30.9346981048584,224 "learning_rate": 9.958522879314958e-07,225 "loss": 0.6242,226 "step": 31227 },228 {229 "epoch": 0.004281509232004281,230 "grad_norm": 15.954740524291992,231 "learning_rate": 9.957184907679957e-07,232 "loss": 0.5834,233 "step": 32234 },235 {236 "epoch": 0.004415306395504416,237 "grad_norm": 42.15216064453125,238 "learning_rate": 9.955846936044955e-07,239 "loss": 0.3828,240 "step": 33241 },242 {243 "epoch": 0.004549103559004549,244 "grad_norm": 24.38614273071289,245 "learning_rate": 9.954508964409954e-07,246 "loss": 0.637,247 "step": 34248 },249 {250 "epoch": 0.004682900722504683,251 "grad_norm": 14.429488182067871,252 "learning_rate": 9.953170992774953e-07,253 "loss": 0.5384,254 "step": 35255 },256 {257 "epoch": 0.004816697886004816,258 "grad_norm": 23.431161880493164,259 "learning_rate": 9.951833021139952e-07,260 "loss": 0.5875,261 "step": 36262 },263 {264 "epoch": 0.0049504950495049506,265 "grad_norm": 23.903627395629883,266 "learning_rate": 9.95049504950495e-07,267 "loss": 0.5366,268 "step": 37269 },270 {271 "epoch": 0.005084292213005084,272 "grad_norm": 30.819042205810547,273 "learning_rate": 9.94915707786995e-07,274 "loss": 0.426,275 "step": 38276 },277 {278 "epoch": 0.005218089376505218,279 "grad_norm": 16.94266700744629,280 "learning_rate": 9.947819106234946e-07,281 "loss": 0.4456,282 "step": 39283 },284 {285 "epoch": 0.005351886540005352,286 "grad_norm": 20.177934646606445,287 "learning_rate": 9.946481134599945e-07,288 "loss": 0.4356,289 "step": 40290 },291 {292 "epoch": 0.0054856837035054855,293 "grad_norm": 63.00517272949219,294 "learning_rate": 9.945143162964946e-07,295 "loss": 0.6703,296 "step": 41297 },298 {299 "epoch": 0.00561948086700562,300 "grad_norm": 44.83641815185547,301 "learning_rate": 9.943805191329945e-07,302 "loss": 0.5356,303 "step": 42304 },305 {306 "epoch": 0.005753278030505753,307 "grad_norm": 46.97093200683594,308 "learning_rate": 9.942467219694941e-07,309 "loss": 0.5233,310 "step": 43311 },312 {313 "epoch": 0.005887075194005887,314 "grad_norm": 27.333566665649414,315 "learning_rate": 9.94112924805994e-07,316 "loss": 0.5151,317 "step": 44318 },319 {320 "epoch": 0.0060208723575060205,321 "grad_norm": 38.97282028198242,322 "learning_rate": 9.939791276424939e-07,323 "loss": 0.7587,324 "step": 45325 },326 {327 "epoch": 0.006154669521006155,328 "grad_norm": 38.45254898071289,329 "learning_rate": 9.938453304789938e-07,330 "loss": 0.4906,331 "step": 46332 },333 {334 "epoch": 0.006288466684506289,335 "grad_norm": 23.768251419067383,336 "learning_rate": 9.937115333154936e-07,337 "loss": 0.7103,338 "step": 47339 },340 {341 "epoch": 0.006422263848006422,342 "grad_norm": 55.38822555541992,343 "learning_rate": 9.935777361519935e-07,344 "loss": 0.7183,345 "step": 48346 },347 {348 "epoch": 0.006556061011506556,349 "grad_norm": 46.217628479003906,350 "learning_rate": 9.934439389884934e-07,351 "loss": 0.5445,352 "step": 49353 },354 {355 "epoch": 0.00668985817500669,356 "grad_norm": 37.99263381958008,357 "learning_rate": 9.933101418249933e-07,358 "loss": 0.4591,359 "step": 50360 },361 {362 "epoch": 0.006823655338506824,363 "grad_norm": 72.31514739990234,364 "learning_rate": 9.931763446614932e-07,365 "loss": 0.4505,366 "step": 51367 },368 {369 "epoch": 0.006957452502006957,370 "grad_norm": 18.87731170654297,371 "learning_rate": 9.93042547497993e-07,372 "loss": 0.5499,373 "step": 52374 },375 {376 "epoch": 0.007091249665507091,377 "grad_norm": 25.828163146972656,378 "learning_rate": 9.92908750334493e-07,379 "loss": 0.376,380 "step": 53381 },382 {383 "epoch": 0.0072250468290072254,384 "grad_norm": 32.08184051513672,385 "learning_rate": 9.927749531709926e-07,386 "loss": 0.5988,387 "step": 54388 },389 {390 "epoch": 0.007358843992507359,391 "grad_norm": 28.96098518371582,392 "learning_rate": 9.926411560074925e-07,393 "loss": 0.5061,394 "step": 55395 },396 {397 "epoch": 0.007492641156007493,398 "grad_norm": 28.637956619262695,399 "learning_rate": 9.925073588439926e-07,400 "loss": 0.4503,401 "step": 56402 },403 {404 "epoch": 0.007626438319507626,405 "grad_norm": 64.53285217285156,406 "learning_rate": 9.923735616804924e-07,407 "loss": 0.7576,408 "step": 57409 },410 {411 "epoch": 0.00776023548300776,412 "grad_norm": 75.2042236328125,413 "learning_rate": 9.92239764516992e-07,414 "loss": 0.9545,415 "step": 58416 },417 {418 "epoch": 0.007894032646507895,419 "grad_norm": 52.87149429321289,420 "learning_rate": 9.92105967353492e-07,421 "loss": 0.7359,422 "step": 59423 },424 {425 "epoch": 0.008027829810008028,426 "grad_norm": 28.44051742553711,427 "learning_rate": 9.919721701899919e-07,428 "loss": 0.5349,429 "step": 60430 },431 {432 "epoch": 0.008161626973508161,433 "grad_norm": 15.476859092712402,434 "learning_rate": 9.918383730264917e-07,435 "loss": 0.5132,436 "step": 61437 },438 {439 "epoch": 0.008295424137008296,440 "grad_norm": 51.01866912841797,441 "learning_rate": 9.917045758629916e-07,442 "loss": 0.3655,443 "step": 62444 },445 {446 "epoch": 0.00842922130050843,447 "grad_norm": 45.629920959472656,448 "learning_rate": 9.915707786994915e-07,449 "loss": 0.5148,450 "step": 63451 },452 {453 "epoch": 0.008563018464008563,454 "grad_norm": 20.70054054260254,455 "learning_rate": 9.914369815359914e-07,456 "loss": 0.5572,457 "step": 64458 },459 {460 "epoch": 0.008696815627508696,461 "grad_norm": 50.97990417480469,462 "learning_rate": 9.913031843724913e-07,463 "loss": 0.4539,464 "step": 65465 },466 {467 "epoch": 0.008830612791008831,468 "grad_norm": 56.28893280029297,469 "learning_rate": 9.911693872089911e-07,470 "loss": 0.5019,471 "step": 66472 },473 {474 "epoch": 0.008964409954508964,475 "grad_norm": 50.34807205200195,476 "learning_rate": 9.91035590045491e-07,477 "loss": 0.4686,478 "step": 67479 },480 {481 "epoch": 0.009098207118009098,482 "grad_norm": 29.993785858154297,483 "learning_rate": 9.90901792881991e-07,484 "loss": 0.6252,485 "step": 68486 },487 {488 "epoch": 0.009232004281509233,489 "grad_norm": 19.4514102935791,490 "learning_rate": 9.907679957184908e-07,491 "loss": 0.3352,492 "step": 69493 },494 {495 "epoch": 0.009365801445009366,496 "grad_norm": 22.55575180053711,497 "learning_rate": 9.906341985549907e-07,498 "loss": 0.5184,499 "step": 70500 },501 {502 "epoch": 0.0094995986085095,503 "grad_norm": 36.30824661254883,504 "learning_rate": 9.905004013914905e-07,505 "loss": 0.5349,506 "step": 71507 },508 {509 "epoch": 0.009633395772009633,510 "grad_norm": 16.004980087280273,511 "learning_rate": 9.903666042279904e-07,512 "loss": 0.2966,513 "step": 72514 },515 {516 "epoch": 0.009767192935509768,517 "grad_norm": 20.882448196411133,518 "learning_rate": 9.902328070644903e-07,519 "loss": 0.3086,520 "step": 73521 },522 {523 "epoch": 0.009900990099009901,524 "grad_norm": 26.98880386352539,525 "learning_rate": 9.9009900990099e-07,526 "loss": 0.5409,527 "step": 74528 },529 {530 "epoch": 0.010034787262510034,531 "grad_norm": 61.62352752685547,532 "learning_rate": 9.899652127374898e-07,533 "loss": 0.789,534 "step": 75535 },536 {537 "epoch": 0.010168584426010168,538 "grad_norm": 42.93553161621094,539 "learning_rate": 9.898314155739897e-07,540 "loss": 0.5571,541 "step": 76542 },543 {544 "epoch": 0.010302381589510303,545 "grad_norm": 84.17617797851562,546 "learning_rate": 9.896976184104896e-07,547 "loss": 0.9816,548 "step": 77549 },550 {551 "epoch": 0.010436178753010436,552 "grad_norm": 25.727781295776367,553 "learning_rate": 9.895638212469895e-07,554 "loss": 0.5069,555 "step": 78556 },557 {558 "epoch": 0.01056997591651057,559 "grad_norm": 32.30039596557617,560 "learning_rate": 9.894300240834894e-07,561 "loss": 0.566,562 "step": 79563 },564 {565 "epoch": 0.010703773080010704,566 "grad_norm": 28.826736450195312,567 "learning_rate": 9.892962269199892e-07,568 "loss": 0.1641,569 "step": 80570 },571 {572 "epoch": 0.010837570243510838,573 "grad_norm": 23.311487197875977,574 "learning_rate": 9.891624297564891e-07,575 "loss": 0.4055,576 "step": 81577 },578 {579 "epoch": 0.010971367407010971,580 "grad_norm": 26.798765182495117,581 "learning_rate": 9.89028632592989e-07,582 "loss": 0.5606,583 "step": 82584 },585 {586 "epoch": 0.011105164570511104,587 "grad_norm": 12.162820816040039,588 "learning_rate": 9.888948354294889e-07,589 "loss": 0.574,590 "step": 83591 },592 {593 "epoch": 0.01123896173401124,594 "grad_norm": 21.961275100708008,595 "learning_rate": 9.887610382659888e-07,596 "loss": 0.5561,597 "step": 84598 },599 {600 "epoch": 0.011372758897511373,601 "grad_norm": 23.370407104492188,602 "learning_rate": 9.886272411024886e-07,603 "loss": 0.3984,604 "step": 85605 },606 {607 "epoch": 0.011506556061011506,608 "grad_norm": 15.90936279296875,609 "learning_rate": 9.884934439389885e-07,610 "loss": 0.6116,611 "step": 86612 },613 {614 "epoch": 0.011640353224511641,615 "grad_norm": 21.821765899658203,616 "learning_rate": 9.883596467754884e-07,617 "loss": 0.5412,618 "step": 87619 },620 {621 "epoch": 0.011774150388011774,622 "grad_norm": 11.008634567260742,623 "learning_rate": 9.882258496119883e-07,624 "loss": 0.6626,625 "step": 88626 },627 {628 "epoch": 0.011907947551511908,629 "grad_norm": 15.92811393737793,630 "learning_rate": 9.88092052448488e-07,631 "loss": 0.525,632 "step": 89633 },634 {635 "epoch": 0.012041744715012041,636 "grad_norm": 20.10358428955078,637 "learning_rate": 9.879582552849878e-07,638 "loss": 0.6513,639 "step": 90640 },641 {642 "epoch": 0.012175541878512176,643 "grad_norm": 16.40944480895996,644 "learning_rate": 9.878244581214877e-07,645 "loss": 0.6041,646 "step": 91647 },648 {649 "epoch": 0.01230933904201231,650 "grad_norm": 10.530786514282227,651 "learning_rate": 9.876906609579878e-07,652 "loss": 0.523,653 "step": 92654 },655 {656 "epoch": 0.012443136205512443,657 "grad_norm": 36.8993034362793,658 "learning_rate": 9.875568637944875e-07,659 "loss": 0.4599,660 "step": 93661 },662 {663 "epoch": 0.012576933369012578,664 "grad_norm": 15.143503189086914,665 "learning_rate": 9.874230666309873e-07,666 "loss": 0.5894,667 "step": 94668 },669 {670 "epoch": 0.012710730532512711,671 "grad_norm": 58.1686897277832,672 "learning_rate": 9.872892694674872e-07,673 "loss": 0.7202,674 "step": 95675 },676 {677 "epoch": 0.012844527696012844,678 "grad_norm": 21.70572853088379,679 "learning_rate": 9.87155472303987e-07,680 "loss": 0.364,681 "step": 96682 },683 {684 "epoch": 0.012978324859512978,685 "grad_norm": 39.490699768066406,686 "learning_rate": 9.87021675140487e-07,687 "loss": 0.6808,688 "step": 97689 },690 {691 "epoch": 0.013112122023013113,692 "grad_norm": 17.027238845825195,693 "learning_rate": 9.868878779769869e-07,694 "loss": 0.5427,695 "step": 98696 },697 {698 "epoch": 0.013245919186513246,699 "grad_norm": 7.477230548858643,700 "learning_rate": 9.867540808134867e-07,701 "loss": 0.4726,702 "step": 99703 },704 {705 "epoch": 0.01337971635001338,706 "grad_norm": 20.803356170654297,707 "learning_rate": 9.866202836499866e-07,708 "loss": 0.5258,709 "step": 100710 },711 {712 "epoch": 0.013513513513513514,713 "grad_norm": 41.74342346191406,714 "learning_rate": 9.864864864864865e-07,715 "loss": 0.6083,716 "step": 101717 },718 {719 "epoch": 0.013647310677013648,720 "grad_norm": 17.93121910095215,721 "learning_rate": 9.863526893229864e-07,722 "loss": 0.4261,723 "step": 102724 },725 {726 "epoch": 0.013781107840513781,727 "grad_norm": 16.341951370239258,728 "learning_rate": 9.862188921594862e-07,729 "loss": 0.548,730 "step": 103731 },732 {733 "epoch": 0.013914905004013914,734 "grad_norm": 15.236637115478516,735 "learning_rate": 9.86085094995986e-07,736 "loss": 0.5264,737 "step": 104738 },739 {740 "epoch": 0.01404870216751405,741 "grad_norm": 29.76473617553711,742 "learning_rate": 9.859512978324858e-07,743 "loss": 0.5176,744 "step": 105745 },746 {747 "epoch": 0.014182499331014183,748 "grad_norm": 17.293731689453125,749 "learning_rate": 9.858175006689857e-07,750 "loss": 0.4871,751 "step": 106752 },753 {754 "epoch": 0.014316296494514316,755 "grad_norm": 19.846126556396484,756 "learning_rate": 9.856837035054858e-07,757 "loss": 0.4828,758 "step": 107759 },760 {761 "epoch": 0.014450093658014451,762 "grad_norm": 32.264102935791016,763 "learning_rate": 9.855499063419854e-07,764 "loss": 0.6576,765 "step": 108766 },767 {768 "epoch": 0.014583890821514584,769 "grad_norm": 36.08279037475586,770 "learning_rate": 9.854161091784853e-07,771 "loss": 0.5862,772 "step": 109773 },774 {775 "epoch": 0.014717687985014717,776 "grad_norm": 15.164156913757324,777 "learning_rate": 9.852823120149852e-07,778 "loss": 0.5147,779 "step": 110780 },781 {782 "epoch": 0.01485148514851485,783 "grad_norm": 44.845733642578125,784 "learning_rate": 9.85148514851485e-07,785 "loss": 0.696,786 "step": 111787 },788 {789 "epoch": 0.014985282312014986,790 "grad_norm": 16.587377548217773,791 "learning_rate": 9.85014717687985e-07,792 "loss": 0.523,793 "step": 112794 },795 {796 "epoch": 0.01511907947551512,797 "grad_norm": 28.191667556762695,798 "learning_rate": 9.848809205244848e-07,799 "loss": 0.343,800 "step": 113801 },802 {803 "epoch": 0.015252876639015252,804 "grad_norm": 18.207569122314453,805 "learning_rate": 9.847471233609847e-07,806 "loss": 0.5568,807 "step": 114808 },809 {810 "epoch": 0.015386673802515387,811 "grad_norm": 12.054566383361816,812 "learning_rate": 9.846133261974846e-07,813 "loss": 0.4707,814 "step": 115815 },816 {817 "epoch": 0.01552047096601552,818 "grad_norm": 14.004857063293457,819 "learning_rate": 9.844795290339845e-07,820 "loss": 0.701,821 "step": 116822 },823 {824 "epoch": 0.015654268129515654,825 "grad_norm": 45.63251876831055,826 "learning_rate": 9.843457318704843e-07,827 "loss": 0.4949,828 "step": 117829 },830 {831 "epoch": 0.01578806529301579,832 "grad_norm": 39.01028060913086,833 "learning_rate": 9.842119347069842e-07,834 "loss": 0.634,835 "step": 118836 },837 {838 "epoch": 0.01592186245651592,839 "grad_norm": 35.620628356933594,840 "learning_rate": 9.84078137543484e-07,841 "loss": 0.4249,842 "step": 119843 },844 {845 "epoch": 0.016055659620016056,846 "grad_norm": 14.283676147460938,847 "learning_rate": 9.839443403799838e-07,848 "loss": 0.5107,849 "step": 120850 },851 {852 "epoch": 0.01618945678351619,853 "grad_norm": 29.358863830566406,854 "learning_rate": 9.838105432164837e-07,855 "loss": 0.6483,856 "step": 121857 },858 {859 "epoch": 0.016323253947016322,860 "grad_norm": 31.07655143737793,861 "learning_rate": 9.836767460529837e-07,862 "loss": 0.6284,863 "step": 122864 },865 {866 "epoch": 0.016457051110516457,867 "grad_norm": 31.927175521850586,868 "learning_rate": 9.835429488894836e-07,869 "loss": 0.3563,870 "step": 123871 },872 {873 "epoch": 0.016590848274016592,874 "grad_norm": 22.668601989746094,875 "learning_rate": 9.834091517259833e-07,876 "loss": 0.4969,877 "step": 124878 },879 {880 "epoch": 0.016724645437516724,881 "grad_norm": 36.73610305786133,882 "learning_rate": 9.832753545624832e-07,883 "loss": 0.7297,884 "step": 125885 },886 {887 "epoch": 0.01685844260101686,888 "grad_norm": 22.381183624267578,889 "learning_rate": 9.83141557398983e-07,890 "loss": 0.5619,891 "step": 126892 },893 {894 "epoch": 0.01699223976451699,895 "grad_norm": 24.720203399658203,896 "learning_rate": 9.83007760235483e-07,897 "loss": 0.5357,898 "step": 127899 },900 {901 "epoch": 0.017126036928017126,902 "grad_norm": 41.036224365234375,903 "learning_rate": 9.828739630719828e-07,904 "loss": 0.6461,905 "step": 128906 },907 {908 "epoch": 0.01725983409151726,909 "grad_norm": 38.702857971191406,910 "learning_rate": 9.827401659084827e-07,911 "loss": 0.4043,912 "step": 129913 },914 {915 "epoch": 0.017393631255017392,916 "grad_norm": 16.72785186767578,917 "learning_rate": 9.826063687449826e-07,918 "loss": 0.6708,919 "step": 130920 },921 {922 "epoch": 0.017527428418517527,923 "grad_norm": 34.862831115722656,924 "learning_rate": 9.824725715814824e-07,925 "loss": 0.4887,926 "step": 131927 },928 {929 "epoch": 0.017661225582017662,930 "grad_norm": 19.050689697265625,931 "learning_rate": 9.823387744179823e-07,932 "loss": 0.4868,933 "step": 132934 },935 {936 "epoch": 0.017795022745517794,937 "grad_norm": 12.097559928894043,938 "learning_rate": 9.822049772544822e-07,939 "loss": 0.463,940 "step": 133941 },942 {943 "epoch": 0.01792881990901793,944 "grad_norm": 18.71462059020996,945 "learning_rate": 9.82071180090982e-07,946 "loss": 0.4137,947 "step": 134948 },949 {950 "epoch": 0.018062617072518064,951 "grad_norm": 42.17513656616211,952 "learning_rate": 9.819373829274818e-07,953 "loss": 0.6429,954 "step": 135955 },956 {957 "epoch": 0.018196414236018196,958 "grad_norm": 13.48039436340332,959 "learning_rate": 9.818035857639818e-07,960 "loss": 0.4156,961 "step": 136962 },963 {964 "epoch": 0.01833021139951833,965 "grad_norm": 49.489158630371094,966 "learning_rate": 9.816697886004817e-07,967 "loss": 0.7119,968 "step": 137969 },970 {971 "epoch": 0.018464008563018466,972 "grad_norm": 22.816030502319336,973 "learning_rate": 9.815359914369816e-07,974 "loss": 0.4514,975 "step": 138976 },977 {978 "epoch": 0.018597805726518597,979 "grad_norm": 10.031615257263184,980 "learning_rate": 9.814021942734813e-07,981 "loss": 0.399,982 "step": 139983 },984 {985 "epoch": 0.018731602890018732,986 "grad_norm": 27.976482391357422,987 "learning_rate": 9.812683971099811e-07,988 "loss": 0.5454,989 "step": 140990 },991 {992 "epoch": 0.018865400053518864,993 "grad_norm": 19.82789421081543,994 "learning_rate": 9.81134599946481e-07,995 "loss": 0.5387,996 "step": 141997 },998 {999 "epoch": 0.018999197217019,1000 "grad_norm": 19.963842391967773,1001 "learning_rate": 9.81000802782981e-07,1002 "loss": 0.4297,1003 "step": 1421004 },1005 {1006 "epoch": 0.019132994380519134,1007 "grad_norm": 12.969765663146973,1008 "learning_rate": 9.808670056194808e-07,1009 "loss": 0.5717,1010 "step": 1431011 },1012 {1013 "epoch": 0.019266791544019266,1014 "grad_norm": 50.279972076416016,1015 "learning_rate": 9.807332084559807e-07,1016 "loss": 0.4063,1017 "step": 1441018 },1019 {1020 "epoch": 0.0194005887075194,1021 "grad_norm": 17.025346755981445,1022 "learning_rate": 9.805994112924805e-07,1023 "loss": 0.489,1024 "step": 1451025 },1026 {1027 "epoch": 0.019534385871019536,1028 "grad_norm": 24.924968719482422,1029 "learning_rate": 9.804656141289804e-07,1030 "loss": 0.4458,1031 "step": 1461032 },1033 {1034 "epoch": 0.019668183034519667,1035 "grad_norm": 35.7073860168457,1036 "learning_rate": 9.803318169654803e-07,1037 "loss": 0.6295,1038 "step": 1471039 },1040 {1041 "epoch": 0.019801980198019802,1042 "grad_norm": 34.52640151977539,1043 "learning_rate": 9.801980198019802e-07,1044 "loss": 0.4543,1045 "step": 1481046 },1047 {1048 "epoch": 0.019935777361519937,1049 "grad_norm": 43.62965774536133,1050 "learning_rate": 9.8006422263848e-07,1051 "loss": 0.6497,1052 "step": 1491053 },1054 {1055 "epoch": 0.02006957452502007,1056 "grad_norm": 8.462423324584961,1057 "learning_rate": 9.7993042547498e-07,1058 "loss": 0.3796,1059 "step": 1501060 },1061 {1062 "epoch": 0.020203371688520204,1063 "grad_norm": 16.452775955200195,1064 "learning_rate": 9.797966283114798e-07,1065 "loss": 0.4846,1066 "step": 1511067 },1068 {1069 "epoch": 0.020337168852020335,1070 "grad_norm": 17.432998657226562,1071 "learning_rate": 9.796628311479797e-07,1072 "loss": 0.5127,1073 "step": 1521074 },1075 {1076 "epoch": 0.02047096601552047,1077 "grad_norm": 26.148113250732422,1078 "learning_rate": 9.795290339844796e-07,1079 "loss": 0.5303,1080 "step": 1531081 },1082 {1083 "epoch": 0.020604763179020606,1084 "grad_norm": 14.63249683380127,1085 "learning_rate": 9.793952368209795e-07,1086 "loss": 0.5467,1087 "step": 1541088 },1089 {1090 "epoch": 0.020738560342520737,1091 "grad_norm": 32.48606491088867,1092 "learning_rate": 9.792614396574791e-07,1093 "loss": 0.6154,1094 "step": 1551095 },1096 {1097 "epoch": 0.020872357506020872,1098 "grad_norm": 36.01746368408203,1099 "learning_rate": 9.79127642493979e-07,1100 "loss": 0.3737,1101 "step": 1561102 },1103 {1104 "epoch": 0.021006154669521007,1105 "grad_norm": 39.24868392944336,1106 "learning_rate": 9.789938453304789e-07,1107 "loss": 0.4689,1108 "step": 1571109 },1110 {1111 "epoch": 0.02113995183302114,1112 "grad_norm": 15.68065357208252,1113 "learning_rate": 9.78860048166979e-07,1114 "loss": 0.4462,1115 "step": 1581116 },1117 {1118 "epoch": 0.021273748996521274,1119 "grad_norm": 40.858089447021484,1120 "learning_rate": 9.787262510034786e-07,1121 "loss": 0.3191,1122 "step": 1591123 },1124 {1125 "epoch": 0.02140754616002141,1126 "grad_norm": 25.862213134765625,1127 "learning_rate": 9.785924538399785e-07,1128 "loss": 0.4464,1129 "step": 1601130 },1131 {1132 "epoch": 0.02154134332352154,1133 "grad_norm": 8.73212718963623,1134 "learning_rate": 9.784586566764784e-07,1135 "loss": 0.4226,1136 "step": 1611137 },1138 {1139 "epoch": 0.021675140487021675,1140 "grad_norm": 30.69873046875,1141 "learning_rate": 9.783248595129783e-07,1142 "loss": 0.5283,1143 "step": 1621144 },1145 {1146 "epoch": 0.02180893765052181,1147 "grad_norm": 24.516950607299805,1148 "learning_rate": 9.781910623494782e-07,1149 "loss": 0.6922,1150 "step": 1631151 },1152 {1153 "epoch": 0.021942734814021942,1154 "grad_norm": 43.645503997802734,1155 "learning_rate": 9.78057265185978e-07,1156 "loss": 0.7168,1157 "step": 1641158 },1159 {1160 "epoch": 0.022076531977522077,1161 "grad_norm": 7.280452251434326,1162 "learning_rate": 9.77923468022478e-07,1163 "loss": 0.3862,1164 "step": 1651165 },1166 {1167 "epoch": 0.02221032914102221,1168 "grad_norm": 20.300607681274414,1169 "learning_rate": 9.777896708589778e-07,1170 "loss": 0.3353,1171 "step": 1661172 },1173 {1174 "epoch": 0.022344126304522344,1175 "grad_norm": 13.733163833618164,1176 "learning_rate": 9.776558736954777e-07,1177 "loss": 0.3542,1178 "step": 1671179 },1180 {1181 "epoch": 0.02247792346802248,1182 "grad_norm": 58.210411071777344,1183 "learning_rate": 9.775220765319776e-07,1184 "loss": 0.7801,1185 "step": 1681186 },1187 {1188 "epoch": 0.02261172063152261,1189 "grad_norm": 31.140893936157227,1190 "learning_rate": 9.773882793684774e-07,1191 "loss": 0.5741,1192 "step": 1691193 },1194 {1195 "epoch": 0.022745517795022745,1196 "grad_norm": 27.565919876098633,1197 "learning_rate": 9.77254482204977e-07,1198 "loss": 0.377,1199 "step": 1701200 },