lightblue/DeepSeek-R1-Distill-Qwen-7B-Multilingual
2219
1{2 "best_metric": null,3 "best_model_checkpoint": null,4 "epoch": 1.0,5 "eval_steps": 11,6 "global_step": 101,7 "is_hyper_param_search": false,8 "is_local_process_zero": true,9 "is_world_process_zero": true,10 "log_history": [11 {12 "epoch": 0.009900990099009901,13 "grad_norm": 1.9865820407867432,14 "learning_rate": 5e-06,15 "loss": 0.8807,16 "step": 117 },18 {19 "epoch": 0.019801980198019802,20 "grad_norm": 1.6995460987091064,21 "learning_rate": 1e-05,22 "loss": 0.7298,23 "step": 224 },25 {26 "epoch": 0.0297029702970297,27 "grad_norm": 1.870273232460022,28 "learning_rate": 9.997482711915926e-06,29 "loss": 0.8256,30 "step": 331 },32 {33 "epoch": 0.039603960396039604,34 "grad_norm": 2.035698175430298,35 "learning_rate": 9.989933382359423e-06,36 "loss": 0.8423,37 "step": 438 },39 {40 "epoch": 0.04950495049504951,41 "grad_norm": 1.5802401304244995,42 "learning_rate": 9.977359612865424e-06,43 "loss": 1.0321,44 "step": 545 },46 {47 "epoch": 0.0594059405940594,48 "grad_norm": 1.3238128423690796,49 "learning_rate": 9.959774064153977e-06,50 "loss": 0.7894,51 "step": 652 },53 {54 "epoch": 0.06930693069306931,55 "grad_norm": 1.3449586629867554,56 "learning_rate": 9.937194443381972e-06,57 "loss": 0.69,58 "step": 759 },60 {61 "epoch": 0.07920792079207921,62 "grad_norm": 1.215697169303894,63 "learning_rate": 9.909643486313533e-06,64 "loss": 0.8263,65 "step": 866 },67 {68 "epoch": 0.0891089108910891,69 "grad_norm": 1.4947011470794678,70 "learning_rate": 9.877148934427037e-06,71 "loss": 0.7706,72 "step": 973 },74 {75 "epoch": 0.09900990099009901,76 "grad_norm": 0.9939355850219727,77 "learning_rate": 9.839743506981783e-06,78 "loss": 0.6854,79 "step": 1080 },81 {82 "epoch": 0.10891089108910891,83 "grad_norm": 1.3050940036773682,84 "learning_rate": 9.797464868072489e-06,85 "loss": 0.9061,86 "step": 1187 },88 {89 "epoch": 0.10891089108910891,90 "eval_loss": 0.7099791765213013,91 "eval_runtime": 1.6876,92 "eval_samples_per_second": 5.333,93 "eval_steps_per_second": 1.185,94 "step": 1195 },96 {97 "epoch": 0.1188118811881188,98 "grad_norm": 1.0027287006378174,99 "learning_rate": 9.750355588704728e-06,100 "loss": 0.7043,101 "step": 12102 },103 {104 "epoch": 0.12871287128712872,105 "grad_norm": 0.9371753334999084,106 "learning_rate": 9.698463103929542e-06,107 "loss": 0.6727,108 "step": 13109 },110 {111 "epoch": 0.13861386138613863,112 "grad_norm": 0.9471062421798706,113 "learning_rate": 9.641839665080363e-06,114 "loss": 0.6772,115 "step": 14116 },117 {118 "epoch": 0.1485148514851485,119 "grad_norm": 1.1542134284973145,120 "learning_rate": 9.580542287160348e-06,121 "loss": 0.7956,122 "step": 15123 },124 {125 "epoch": 0.15841584158415842,126 "grad_norm": 1.1275238990783691,127 "learning_rate": 9.514632691433108e-06,128 "loss": 0.7486,129 "step": 16130 },131 {132 "epoch": 0.16831683168316833,133 "grad_norm": 1.0453617572784424,134 "learning_rate": 9.444177243274619e-06,135 "loss": 0.659,136 "step": 17137 },138 {139 "epoch": 0.1782178217821782,140 "grad_norm": 1.0331149101257324,141 "learning_rate": 9.369246885348926e-06,142 "loss": 0.7471,143 "step": 18144 },145 {146 "epoch": 0.18811881188118812,147 "grad_norm": 0.8706634044647217,148 "learning_rate": 9.289917066174887e-06,149 "loss": 0.5547,150 "step": 19151 },152 {153 "epoch": 0.19801980198019803,154 "grad_norm": 1.1781532764434814,155 "learning_rate": 9.206267664155906e-06,156 "loss": 0.6239,157 "step": 20158 },159 {160 "epoch": 0.2079207920792079,161 "grad_norm": 1.2011443376541138,162 "learning_rate": 9.118382907149164e-06,163 "loss": 0.7847,164 "step": 21165 },166 {167 "epoch": 0.21782178217821782,168 "grad_norm": 1.028536319732666,169 "learning_rate": 9.026351287655294e-06,170 "loss": 0.6565,171 "step": 22172 },173 {174 "epoch": 0.21782178217821782,175 "eval_loss": 0.6684643030166626,176 "eval_runtime": 1.6848,177 "eval_samples_per_second": 5.342,178 "eval_steps_per_second": 1.187,179 "step": 22180 },181 {182 "epoch": 0.22772277227722773,183 "grad_norm": 0.9103761315345764,184 "learning_rate": 8.930265473713939e-06,185 "loss": 0.6495,186 "step": 23187 },188 {189 "epoch": 0.2376237623762376,190 "grad_norm": 1.0075730085372925,191 "learning_rate": 8.83022221559489e-06,192 "loss": 0.7109,193 "step": 24194 },195 {196 "epoch": 0.24752475247524752,197 "grad_norm": 0.9617104530334473,198 "learning_rate": 8.726322248378775e-06,199 "loss": 0.6993,200 "step": 25201 },202 {203 "epoch": 0.25742574257425743,204 "grad_norm": 1.072245717048645,205 "learning_rate": 8.61867019052535e-06,206 "loss": 0.8897,207 "step": 26208 },209 {210 "epoch": 0.26732673267326734,211 "grad_norm": 1.0244859457015991,212 "learning_rate": 8.507374438531606e-06,213 "loss": 0.8158,214 "step": 27215 },216 {217 "epoch": 0.27722772277227725,218 "grad_norm": 1.0862797498703003,219 "learning_rate": 8.392547057785662e-06,220 "loss": 0.7539,221 "step": 28222 },223 {224 "epoch": 0.2871287128712871,225 "grad_norm": 0.9954370260238647,226 "learning_rate": 8.274303669726427e-06,227 "loss": 0.718,228 "step": 29229 },230 {231 "epoch": 0.297029702970297,232 "grad_norm": 1.0018794536590576,233 "learning_rate": 8.152763335422612e-06,234 "loss": 0.7518,235 "step": 30236 },237 {238 "epoch": 0.3069306930693069,239 "grad_norm": 0.9397566318511963,240 "learning_rate": 8.028048435688333e-06,241 "loss": 0.663,242 "step": 31243 },244 {245 "epoch": 0.31683168316831684,246 "grad_norm": 0.8733705878257751,247 "learning_rate": 7.900284547855992e-06,248 "loss": 0.6565,249 "step": 32250 },251 {252 "epoch": 0.32673267326732675,253 "grad_norm": 0.938222348690033,254 "learning_rate": 7.769600319330553e-06,255 "loss": 0.601,256 "step": 33257 },258 {259 "epoch": 0.32673267326732675,260 "eval_loss": 0.648015022277832,261 "eval_runtime": 1.6839,262 "eval_samples_per_second": 5.345,263 "eval_steps_per_second": 1.188,264 "step": 33265 },266 {267 "epoch": 0.33663366336633666,268 "grad_norm": 0.9610555768013,269 "learning_rate": 7.636127338052513e-06,270 "loss": 0.7488,271 "step": 34272 },273 {274 "epoch": 0.3465346534653465,275 "grad_norm": 1.0051265954971313,276 "learning_rate": 7.500000000000001e-06,277 "loss": 0.7929,278 "step": 35279 },280 {281 "epoch": 0.3564356435643564,282 "grad_norm": 0.8689800500869751,283 "learning_rate": 7.361355373863415e-06,284 "loss": 0.57,285 "step": 36286 },287 {288 "epoch": 0.36633663366336633,289 "grad_norm": 1.0085370540618896,290 "learning_rate": 7.2203330630288714e-06,291 "loss": 0.672,292 "step": 37293 },294 {295 "epoch": 0.37623762376237624,296 "grad_norm": 0.8325203061103821,297 "learning_rate": 7.0770750650094335e-06,298 "loss": 0.6143,299 "step": 38300 },301 {302 "epoch": 0.38613861386138615,303 "grad_norm": 0.9245986342430115,304 "learning_rate": 6.931725628465643e-06,305 "loss": 0.7288,306 "step": 39307 },308 {309 "epoch": 0.39603960396039606,310 "grad_norm": 1.1846799850463867,311 "learning_rate": 6.78443110795936e-06,312 "loss": 0.8207,313 "step": 40314 },315 {316 "epoch": 0.40594059405940597,317 "grad_norm": 0.8949479460716248,318 "learning_rate": 6.635339816587109e-06,319 "loss": 0.6476,320 "step": 41321 },322 {323 "epoch": 0.4158415841584158,324 "grad_norm": 0.9853402376174927,325 "learning_rate": 6.484601876641375e-06,326 "loss": 0.7146,327 "step": 42328 },329 {330 "epoch": 0.42574257425742573,331 "grad_norm": 0.9564022421836853,332 "learning_rate": 6.332369068450175e-06,333 "loss": 0.7382,334 "step": 43335 },336 {337 "epoch": 0.43564356435643564,338 "grad_norm": 1.0044441223144531,339 "learning_rate": 6.178794677547138e-06,340 "loss": 0.7668,341 "step": 44342 },343 {344 "epoch": 0.43564356435643564,345 "eval_loss": 0.6342881917953491,346 "eval_runtime": 1.6847,347 "eval_samples_per_second": 5.342,348 "eval_steps_per_second": 1.187,349 "step": 44350 },351 {352 "epoch": 0.44554455445544555,353 "grad_norm": 0.8848042488098145,354 "learning_rate": 6.024033340325954e-06,355 "loss": 0.6532,356 "step": 45357 },358 {359 "epoch": 0.45544554455445546,360 "grad_norm": 0.8762900233268738,361 "learning_rate": 5.8682408883346535e-06,362 "loss": 0.674,363 "step": 46364 },365 {366 "epoch": 0.46534653465346537,367 "grad_norm": 0.8674903512001038,368 "learning_rate": 5.711574191366427e-06,369 "loss": 0.6138,370 "step": 47371 },372 {373 "epoch": 0.4752475247524752,374 "grad_norm": 0.9267687201499939,375 "learning_rate": 5.5541909995050554e-06,376 "loss": 0.6838,377 "step": 48378 },379 {380 "epoch": 0.48514851485148514,381 "grad_norm": 0.8423399329185486,382 "learning_rate": 5.396249784283943e-06,383 "loss": 0.6081,384 "step": 49385 },386 {387 "epoch": 0.49504950495049505,388 "grad_norm": 0.9809019565582275,389 "learning_rate": 5.237909579118713e-06,390 "loss": 0.7202,391 "step": 50392 },393 {394 "epoch": 0.504950495049505,395 "grad_norm": 0.8688491582870483,396 "learning_rate": 5.07932981917404e-06,397 "loss": 0.6236,398 "step": 51399 },400 {401 "epoch": 0.5148514851485149,402 "grad_norm": 0.9908064007759094,403 "learning_rate": 4.9206701808259605e-06,404 "loss": 0.7449,405 "step": 52406 },407 {408 "epoch": 0.5247524752475248,409 "grad_norm": 0.935342013835907,410 "learning_rate": 4.762090420881289e-06,411 "loss": 0.6582,412 "step": 53413 },414 {415 "epoch": 0.5346534653465347,416 "grad_norm": 1.0338435173034668,417 "learning_rate": 4.603750215716057e-06,418 "loss": 0.7808,419 "step": 54420 },421 {422 "epoch": 0.5445544554455446,423 "grad_norm": 1.0253307819366455,424 "learning_rate": 4.445809000494945e-06,425 "loss": 0.8058,426 "step": 55427 },428 {429 "epoch": 0.5445544554455446,430 "eval_loss": 0.6248189210891724,431 "eval_runtime": 1.6835,432 "eval_samples_per_second": 5.346,433 "eval_steps_per_second": 1.188,434 "step": 55435 },436 {437 "epoch": 0.5544554455445545,438 "grad_norm": 0.9448875188827515,439 "learning_rate": 4.2884258086335755e-06,440 "loss": 0.7117,441 "step": 56442 },443 {444 "epoch": 0.5643564356435643,445 "grad_norm": 0.8737557530403137,446 "learning_rate": 4.131759111665349e-06,447 "loss": 0.5988,448 "step": 57449 },450 {451 "epoch": 0.5742574257425742,452 "grad_norm": 0.764798104763031,453 "learning_rate": 3.975966659674048e-06,454 "loss": 0.501,455 "step": 58456 },457 {458 "epoch": 0.5841584158415841,459 "grad_norm": 0.9195338487625122,460 "learning_rate": 3.821205322452863e-06,461 "loss": 0.677,462 "step": 59463 },464 {465 "epoch": 0.594059405940594,466 "grad_norm": 0.9102662801742554,467 "learning_rate": 3.667630931549826e-06,468 "loss": 0.6363,469 "step": 60470 },471 {472 "epoch": 0.6039603960396039,473 "grad_norm": 1.1512144804000854,474 "learning_rate": 3.5153981233586277e-06,475 "loss": 0.7082,476 "step": 61477 },478 {479 "epoch": 0.6138613861386139,480 "grad_norm": 1.0006070137023926,481 "learning_rate": 3.3646601834128924e-06,482 "loss": 0.6685,483 "step": 62484 },485 {486 "epoch": 0.6237623762376238,487 "grad_norm": 0.9273307919502258,488 "learning_rate": 3.2155688920406415e-06,489 "loss": 0.6652,490 "step": 63491 },492 {493 "epoch": 0.6336633663366337,494 "grad_norm": 0.9092214703559875,495 "learning_rate": 3.0682743715343565e-06,496 "loss": 0.6497,497 "step": 64498 },499 {500 "epoch": 0.6435643564356436,501 "grad_norm": 0.8190207481384277,502 "learning_rate": 2.9229249349905686e-06,503 "loss": 0.5194,504 "step": 65505 },506 {507 "epoch": 0.6534653465346535,508 "grad_norm": 0.9039291739463806,509 "learning_rate": 2.7796669369711294e-06,510 "loss": 0.675,511 "step": 66512 },513 {514 "epoch": 0.6534653465346535,515 "eval_loss": 0.6190232038497925,516 "eval_runtime": 1.6846,517 "eval_samples_per_second": 5.342,518 "eval_steps_per_second": 1.187,519 "step": 66520 },521 {522 "epoch": 0.6633663366336634,523 "grad_norm": 1.0311757326126099,524 "learning_rate": 2.6386446261365874e-06,525 "loss": 0.8244,526 "step": 67527 },528 {529 "epoch": 0.6732673267326733,530 "grad_norm": 0.8113132119178772,531 "learning_rate": 2.5000000000000015e-06,532 "loss": 0.5488,533 "step": 68534 },535 {536 "epoch": 0.6831683168316832,537 "grad_norm": 0.9003159403800964,538 "learning_rate": 2.363872661947488e-06,539 "loss": 0.7291,540 "step": 69541 },542 {543 "epoch": 0.693069306930693,544 "grad_norm": 1.057133674621582,545 "learning_rate": 2.230399680669449e-06,546 "loss": 0.8625,547 "step": 70548 },549 {550 "epoch": 0.7029702970297029,551 "grad_norm": 0.9384030699729919,552 "learning_rate": 2.09971545214401e-06,553 "loss": 0.6813,554 "step": 71555 },556 {557 "epoch": 0.7128712871287128,558 "grad_norm": 0.8832238912582397,559 "learning_rate": 1.971951564311668e-06,560 "loss": 0.6099,561 "step": 72562 },563 {564 "epoch": 0.7227722772277227,565 "grad_norm": 0.8297242522239685,566 "learning_rate": 1.8472366645773892e-06,567 "loss": 0.5465,568 "step": 73569 },570 {571 "epoch": 0.7326732673267327,572 "grad_norm": 0.9183589816093445,573 "learning_rate": 1.7256963302735752e-06,574 "loss": 0.6971,575 "step": 74576 },577 {578 "epoch": 0.7425742574257426,579 "grad_norm": 0.8353100419044495,580 "learning_rate": 1.6074529422143398e-06,581 "loss": 0.5426,582 "step": 75583 },584 {585 "epoch": 0.7524752475247525,586 "grad_norm": 1.051756739616394,587 "learning_rate": 1.4926255614683931e-06,588 "loss": 0.7567,589 "step": 76590 },591 {592 "epoch": 0.7623762376237624,593 "grad_norm": 0.8063908815383911,594 "learning_rate": 1.3813298094746491e-06,595 "loss": 0.5383,596 "step": 77597 },598 {599 "epoch": 0.7623762376237624,600 "eval_loss": 0.6155756711959839,601 "eval_runtime": 1.6854,602 "eval_samples_per_second": 5.34,603 "eval_steps_per_second": 1.187,604 "step": 77605 },606 {607 "epoch": 0.7722772277227723,608 "grad_norm": 0.8472511768341064,609 "learning_rate": 1.2736777516212267e-06,610 "loss": 0.5792,611 "step": 78612 },613 {614 "epoch": 0.7821782178217822,615 "grad_norm": 0.8465410470962524,616 "learning_rate": 1.1697777844051105e-06,617 "loss": 0.6283,618 "step": 79619 },620 {621 "epoch": 0.7920792079207921,622 "grad_norm": 0.7360741496086121,623 "learning_rate": 1.0697345262860638e-06,624 "loss": 0.4699,625 "step": 80626 },627 {628 "epoch": 0.801980198019802,629 "grad_norm": 0.9224135875701904,630 "learning_rate": 9.73648712344707e-07,631 "loss": 0.6806,632 "step": 81633 },634 {635 "epoch": 0.8118811881188119,636 "grad_norm": 0.871505081653595,637 "learning_rate": 8.816170928508367e-07,638 "loss": 0.6824,639 "step": 82640 },641 {642 "epoch": 0.8217821782178217,643 "grad_norm": 0.8890396952629089,644 "learning_rate": 7.937323358440935e-07,645 "loss": 0.6427,646 "step": 83647 },648 {649 "epoch": 0.8316831683168316,650 "grad_norm": 1.091757893562317,651 "learning_rate": 7.100829338251147e-07,652 "loss": 0.7535,653 "step": 84654 },655 {656 "epoch": 0.8415841584158416,657 "grad_norm": 0.809833824634552,658 "learning_rate": 6.307531146510754e-07,659 "loss": 0.5965,660 "step": 85661 },662 {663 "epoch": 0.8514851485148515,664 "grad_norm": 0.861588716506958,665 "learning_rate": 5.558227567253832e-07,666 "loss": 0.6629,667 "step": 86668 },669 {670 "epoch": 0.8613861386138614,671 "grad_norm": 0.7921182513237,672 "learning_rate": 4.853673085668947e-07,673 "loss": 0.482,674 "step": 87675 },676 {677 "epoch": 0.8712871287128713,678 "grad_norm": 0.7692471742630005,679 "learning_rate": 4.194577128396521e-07,680 "loss": 0.5044,681 "step": 88682 },683 {684 "epoch": 0.8712871287128713,685 "eval_loss": 0.6139124035835266,686 "eval_runtime": 1.684,687 "eval_samples_per_second": 5.344,688 "eval_steps_per_second": 1.188,689 "step": 88690 },691 {692 "epoch": 0.8811881188118812,693 "grad_norm": 0.8603846430778503,694 "learning_rate": 3.581603349196372e-07,695 "loss": 0.6439,696 "step": 89697 },698 {699 "epoch": 0.8910891089108911,700 "grad_norm": 0.8953405022621155,701 "learning_rate": 3.015368960704584e-07,702 "loss": 0.6881,703 "step": 90704 },705 {706 "epoch": 0.900990099009901,707 "grad_norm": 0.9108988046646118,708 "learning_rate": 2.4964441129527337e-07,709 "loss": 0.7118,710 "step": 91711 },712 {713 "epoch": 0.9108910891089109,714 "grad_norm": 0.886647641658783,715 "learning_rate": 2.0253513192751374e-07,716 "loss": 0.7544,717 "step": 92718 },719 {720 "epoch": 0.9207920792079208,721 "grad_norm": 0.943878173828125,722 "learning_rate": 1.6025649301821877e-07,723 "loss": 0.7527,724 "step": 93725 },726 {727 "epoch": 0.9306930693069307,728 "grad_norm": 0.8354966044425964,729 "learning_rate": 1.2285106557296479e-07,730 "loss": 0.6079,731 "step": 94732 },733 {734 "epoch": 0.9405940594059405,735 "grad_norm": 0.8121664524078369,736 "learning_rate": 9.035651368646647e-08,737 "loss": 0.5891,738 "step": 95739 },740 {741 "epoch": 0.9504950495049505,742 "grad_norm": 0.9194057583808899,743 "learning_rate": 6.280555661802857e-08,744 "loss": 0.6953,745 "step": 96746 },747 {748 "epoch": 0.9603960396039604,749 "grad_norm": 0.9447068572044373,750 "learning_rate": 4.02259358460233e-08,751 "loss": 0.7401,752 "step": 97753 },754 {755 "epoch": 0.9702970297029703,756 "grad_norm": 0.8672422766685486,757 "learning_rate": 2.264038713457706e-08,758 "loss": 0.6349,759 "step": 98760 },761 {762 "epoch": 0.9801980198019802,763 "grad_norm": 0.904771625995636,764 "learning_rate": 1.006661764057837e-08,765 "loss": 0.6322,766 "step": 99767 },768 {769 "epoch": 0.9801980198019802,770 "eval_loss": 0.6136024594306946,771 "eval_runtime": 1.6864,772 "eval_samples_per_second": 5.337,773 "eval_steps_per_second": 1.186,774 "step": 99775 },776 {777 "epoch": 0.9900990099009901,778 "grad_norm": 0.8689496517181396,779 "learning_rate": 2.5172880840745873e-09,780 "loss": 0.6191,781 "step": 100782 },783 {784 "epoch": 1.0,785 "grad_norm": 0.8673611283302307,786 "learning_rate": 0.0,787 "loss": 0.5897,788 "step": 101789 },790 {791 "epoch": 1.0,792 "step": 101,793 "total_flos": 1.5438974547499418e+17,794 "train_loss": 0.6889385657735391,795 "train_runtime": 730.4137,796 "train_samples_per_second": 1.101,797 "train_steps_per_second": 0.138798 }799 ],800 "logging_steps": 1,801 "max_steps": 101,802 "num_input_tokens_seen": 0,803 "num_train_epochs": 1,804 "save_steps": 101,805 "stateful_callbacks": {806 "TrainerControl": {807 "args": {808 "should_epoch_stop": false,809 "should_evaluate": false,810 "should_log": false,811 "should_save": true,812 "should_training_stop": true813 },814 "attributes": {}815 }816 },817 "total_flos": 1.5438974547499418e+17,818 "train_batch_size": 1,819 "trial_name": null,820 "trial_params": null821}822