lime-nlp/ExeVRM-8B
529
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 1.0,6 "eval_steps": 200,7 "global_step": 3369,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.00029682398337785694,14 "grad_norm": 16.833446502685547,15 "learning_rate": 1.483679525222552e-08,16 "loss": 0.1938,17 "step": 1,18 "video_reward_cumulative_accuracy": 0.519 },20 {21 "epoch": 0.0005936479667557139,22 "grad_norm": 24.454694747924805,23 "learning_rate": 2.967359050445104e-08,24 "loss": 0.3887,25 "step": 2,26 "video_reward_cumulative_accuracy": 0.527 },28 {29 "epoch": 0.0008904719501335708,30 "grad_norm": 21.62911605834961,31 "learning_rate": 4.451038575667656e-08,32 "loss": 0.2016,33 "step": 3,34 "video_reward_cumulative_accuracy": 0.535 },36 {37 "epoch": 0.0011872959335114278,38 "grad_norm": 18.787561416625977,39 "learning_rate": 5.934718100890208e-08,40 "loss": 0.1834,41 "step": 4,42 "video_reward_cumulative_accuracy": 0.62543 },44 {45 "epoch": 0.0014841199168892847,46 "grad_norm": 25.317777633666992,47 "learning_rate": 7.418397626112761e-08,48 "loss": 0.3673,49 "step": 5,50 "video_reward_cumulative_accuracy": 0.751 },52 {53 "epoch": 0.0017809439002671415,54 "grad_norm": 12.732484817504883,55 "learning_rate": 8.902077151335312e-08,56 "loss": 0.1034,57 "step": 6,58 "video_reward_cumulative_accuracy": 0.583333333333333459 },60 {61 "epoch": 0.0020777678836449986,62 "grad_norm": 20.14723777770996,63 "learning_rate": 1.0385756676557864e-07,64 "loss": 0.2094,65 "step": 7,66 "video_reward_cumulative_accuracy": 0.571428571428571467 },68 {69 "epoch": 0.0023745918670228555,70 "grad_norm": 33.73411178588867,71 "learning_rate": 1.1869436201780416e-07,72 "loss": 0.3328,73 "step": 8,74 "video_reward_cumulative_accuracy": 0.562575 },76 {77 "epoch": 0.0026714158504007124,78 "grad_norm": 21.074481964111328,79 "learning_rate": 1.3353115727002968e-07,80 "loss": 0.1414,81 "step": 9,82 "video_reward_cumulative_accuracy": 0.555555555555555683 },84 {85 "epoch": 0.0029682398337785693,86 "grad_norm": 23.00200080871582,87 "learning_rate": 1.4836795252225522e-07,88 "loss": 0.4463,89 "step": 10,90 "video_reward_cumulative_accuracy": 0.5591 },92 {93 "epoch": 0.003265063817156426,94 "grad_norm": 23.951406478881836,95 "learning_rate": 1.6320474777448073e-07,96 "loss": 0.3231,97 "step": 11,98 "video_reward_cumulative_accuracy": 0.590909090909090999 },100 {101 "epoch": 0.003561887800534283,102 "grad_norm": 25.632526397705078,103 "learning_rate": 1.7804154302670624e-07,104 "loss": 0.2746,105 "step": 12,106 "video_reward_cumulative_accuracy": 0.625107 },108 {109 "epoch": 0.00385871178391214,110 "grad_norm": 17.966325759887695,111 "learning_rate": 1.9287833827893176e-07,112 "loss": 0.224,113 "step": 13,114 "video_reward_cumulative_accuracy": 0.6538461538461539115 },116 {117 "epoch": 0.004155535767289997,118 "grad_norm": 27.712692260742188,119 "learning_rate": 2.0771513353115727e-07,120 "loss": 0.1764,121 "step": 14,122 "video_reward_cumulative_accuracy": 0.6071428571428571123 },124 {125 "epoch": 0.004452359750667854,126 "grad_norm": 24.923372268676758,127 "learning_rate": 2.225519287833828e-07,128 "loss": 0.2788,129 "step": 15,130 "video_reward_cumulative_accuracy": 0.6333333333333333131 },132 {133 "epoch": 0.004749183734045711,134 "grad_norm": 13.765485763549805,135 "learning_rate": 2.3738872403560833e-07,136 "loss": 0.1745,137 "step": 16,138 "video_reward_cumulative_accuracy": 0.65625139 },140 {141 "epoch": 0.0050460077174235675,142 "grad_norm": 17.97304344177246,143 "learning_rate": 2.5222551928783384e-07,144 "loss": 0.2362,145 "step": 17,146 "video_reward_cumulative_accuracy": 0.6470588235294118147 },148 {149 "epoch": 0.005342831700801425,150 "grad_norm": 10.485559463500977,151 "learning_rate": 2.6706231454005935e-07,152 "loss": 0.1295,153 "step": 18,154 "video_reward_cumulative_accuracy": 0.6666666666666666155 },156 {157 "epoch": 0.005639655684179281,158 "grad_norm": 15.357043266296387,159 "learning_rate": 2.8189910979228487e-07,160 "loss": 0.2338,161 "step": 19,162 "video_reward_cumulative_accuracy": 0.6842105263157895163 },164 {165 "epoch": 0.005936479667557139,166 "grad_norm": 15.526144027709961,167 "learning_rate": 2.9673590504451043e-07,168 "loss": 0.2443,169 "step": 20,170 "video_reward_cumulative_accuracy": 0.7171 },172 {173 "epoch": 0.006233303650934996,174 "grad_norm": 19.776208877563477,175 "learning_rate": 3.1157270029673595e-07,176 "loss": 0.1632,177 "step": 21,178 "video_reward_cumulative_accuracy": 0.6904761904761905179 },180 {181 "epoch": 0.006530127634312852,182 "grad_norm": 31.176301956176758,183 "learning_rate": 3.2640949554896146e-07,184 "loss": 0.3122,185 "step": 22,186 "video_reward_cumulative_accuracy": 0.6590909090909091187 },188 {189 "epoch": 0.00682695161769071,190 "grad_norm": 12.820406913757324,191 "learning_rate": 3.41246290801187e-07,192 "loss": 0.1422,193 "step": 23,194 "video_reward_cumulative_accuracy": 0.6521739130434783195 },196 {197 "epoch": 0.007123775601068566,198 "grad_norm": 12.706981658935547,199 "learning_rate": 3.560830860534125e-07,200 "loss": 0.169,201 "step": 24,202 "video_reward_cumulative_accuracy": 0.6458333333333334203 },204 {205 "epoch": 0.0074205995844464235,206 "grad_norm": 16.90671157836914,207 "learning_rate": 3.70919881305638e-07,208 "loss": 0.1716,209 "step": 25,210 "video_reward_cumulative_accuracy": 0.64211 },212 {213 "epoch": 0.00771742356782428,214 "grad_norm": 12.562002182006836,215 "learning_rate": 3.857566765578635e-07,216 "loss": 0.1565,217 "step": 26,218 "video_reward_cumulative_accuracy": 0.6538461538461539219 },220 {221 "epoch": 0.008014247551202136,222 "grad_norm": 10.192179679870605,223 "learning_rate": 4.005934718100891e-07,224 "loss": 0.1055,225 "step": 27,226 "video_reward_cumulative_accuracy": 0.6481481481481481227 },228 {229 "epoch": 0.008311071534579995,230 "grad_norm": 15.415340423583984,231 "learning_rate": 4.1543026706231454e-07,232 "loss": 0.2231,233 "step": 28,234 "video_reward_cumulative_accuracy": 0.6428571428571429235 },236 {237 "epoch": 0.008607895517957851,238 "grad_norm": 21.16411781311035,239 "learning_rate": 4.302670623145401e-07,240 "loss": 0.3278,241 "step": 29,242 "video_reward_cumulative_accuracy": 0.6379310344827587243 },244 {245 "epoch": 0.008904719501335707,246 "grad_norm": 21.195640563964844,247 "learning_rate": 4.451038575667656e-07,248 "loss": 0.2429,249 "step": 30,250 "video_reward_cumulative_accuracy": 0.6333333333333333251 },252 {253 "epoch": 0.009201543484713566,254 "grad_norm": 10.36475658416748,255 "learning_rate": 4.5994065281899114e-07,256 "loss": 0.125,257 "step": 31,258 "video_reward_cumulative_accuracy": 0.6290322580645161259 },260 {261 "epoch": 0.009498367468091422,262 "grad_norm": 14.530949592590332,263 "learning_rate": 4.7477744807121665e-07,264 "loss": 0.2392,265 "step": 32,266 "video_reward_cumulative_accuracy": 0.640625267 },268 {269 "epoch": 0.009795191451469279,270 "grad_norm": 17.089160919189453,271 "learning_rate": 4.896142433234421e-07,272 "loss": 0.2253,273 "step": 33,274 "video_reward_cumulative_accuracy": 0.6363636363636364275 },276 {277 "epoch": 0.010092015434847135,278 "grad_norm": 14.45820426940918,279 "learning_rate": 5.044510385756677e-07,280 "loss": 0.2298,281 "step": 34,282 "video_reward_cumulative_accuracy": 0.6323529411764706283 },284 {285 "epoch": 0.010388839418224993,286 "grad_norm": 12.284329414367676,287 "learning_rate": 5.192878338278932e-07,288 "loss": 0.0988,289 "step": 35,290 "video_reward_cumulative_accuracy": 0.6428571428571429291 },292 {293 "epoch": 0.01068566340160285,294 "grad_norm": 14.295129776000977,295 "learning_rate": 5.341246290801187e-07,296 "loss": 0.2106,297 "step": 36,298 "video_reward_cumulative_accuracy": 0.6388888888888888299 },300 {301 "epoch": 0.010982487384980706,302 "grad_norm": 7.684972286224365,303 "learning_rate": 5.489614243323443e-07,304 "loss": 0.0939,305 "step": 37,306 "video_reward_cumulative_accuracy": 0.6486486486486487307 },308 {309 "epoch": 0.011279311368358563,310 "grad_norm": 9.849855422973633,311 "learning_rate": 5.637982195845697e-07,312 "loss": 0.1303,313 "step": 38,314 "video_reward_cumulative_accuracy": 0.6578947368421053315 },316 {317 "epoch": 0.01157613535173642,318 "grad_norm": 10.725906372070312,319 "learning_rate": 5.786350148367953e-07,320 "loss": 0.1924,321 "step": 39,322 "video_reward_cumulative_accuracy": 0.6538461538461539323 },324 {325 "epoch": 0.011872959335114277,326 "grad_norm": 9.24576473236084,327 "learning_rate": 5.934718100890209e-07,328 "loss": 0.108,329 "step": 40,330 "video_reward_cumulative_accuracy": 0.6375331 },332 {333 "epoch": 0.012169783318492134,334 "grad_norm": 6.677659511566162,335 "learning_rate": 6.083086053412463e-07,336 "loss": 0.0879,337 "step": 41,338 "video_reward_cumulative_accuracy": 0.6463414634146342339 },340 {341 "epoch": 0.012466607301869992,342 "grad_norm": 7.723426342010498,343 "learning_rate": 6.231454005934719e-07,344 "loss": 0.073,345 "step": 42,346 "video_reward_cumulative_accuracy": 0.6428571428571429347 },348 {349 "epoch": 0.012763431285247848,350 "grad_norm": 10.541435241699219,351 "learning_rate": 6.379821958456974e-07,352 "loss": 0.1862,353 "step": 43,354 "video_reward_cumulative_accuracy": 0.6395348837209303355 },356 {357 "epoch": 0.013060255268625705,358 "grad_norm": 7.638758659362793,359 "learning_rate": 6.528189910979229e-07,360 "loss": 0.1184,361 "step": 44,362 "video_reward_cumulative_accuracy": 0.6477272727272727363 },364 {365 "epoch": 0.013357079252003561,366 "grad_norm": 10.385604858398438,367 "learning_rate": 6.676557863501485e-07,368 "loss": 0.1647,369 "step": 45,370 "video_reward_cumulative_accuracy": 0.6555555555555556371 },372 {373 "epoch": 0.01365390323538142,374 "grad_norm": 6.365200996398926,375 "learning_rate": 6.82492581602374e-07,376 "loss": 0.0985,377 "step": 46,378 "video_reward_cumulative_accuracy": 0.6413043478260869379 },380 {381 "epoch": 0.013950727218759276,382 "grad_norm": 7.856001853942871,383 "learning_rate": 6.973293768545995e-07,384 "loss": 0.1468,385 "step": 47,386 "video_reward_cumulative_accuracy": 0.6382978723404256387 },388 {389 "epoch": 0.014247551202137132,390 "grad_norm": 6.851215839385986,391 "learning_rate": 7.12166172106825e-07,392 "loss": 0.0869,393 "step": 48,394 "video_reward_cumulative_accuracy": 0.6354166666666666395 },396 {397 "epoch": 0.014544375185514989,398 "grad_norm": 7.089118003845215,399 "learning_rate": 7.270029673590504e-07,400 "loss": 0.1313,401 "step": 49,402 "video_reward_cumulative_accuracy": 0.6326530612244898403 },404 {405 "epoch": 0.014841199168892847,406 "grad_norm": 3.806837558746338,407 "learning_rate": 7.41839762611276e-07,408 "loss": 0.0955,409 "step": 50,410 "video_reward_cumulative_accuracy": 0.63411 },412 {413 "epoch": 0.015138023152270703,414 "grad_norm": 3.002065420150757,415 "learning_rate": 7.566765578635016e-07,416 "loss": 0.0829,417 "step": 51,418 "video_reward_cumulative_accuracy": 0.6372549019607843419 },420 {421 "epoch": 0.01543484713564856,422 "grad_norm": 8.006941795349121,423 "learning_rate": 7.71513353115727e-07,424 "loss": 0.1208,425 "step": 52,426 "video_reward_cumulative_accuracy": 0.6442307692307693427 },428 {429 "epoch": 0.015731671119026416,430 "grad_norm": 4.04028844833374,431 "learning_rate": 7.863501483679525e-07,432 "loss": 0.0905,433 "step": 53,434 "video_reward_cumulative_accuracy": 0.6415094339622641435 },436 {437 "epoch": 0.016028495102404273,438 "grad_norm": 11.572624206542969,439 "learning_rate": 8.011869436201782e-07,440 "loss": 0.1112,441 "step": 54,442 "video_reward_cumulative_accuracy": 0.6296296296296297443 },444 {445 "epoch": 0.016325319085782133,446 "grad_norm": 2.590923309326172,447 "learning_rate": 8.160237388724036e-07,448 "loss": 0.0886,449 "step": 55,450 "video_reward_cumulative_accuracy": 0.6363636363636364451 },452 {453 "epoch": 0.01662214306915999,454 "grad_norm": 6.2428717613220215,455 "learning_rate": 8.308605341246291e-07,456 "loss": 0.1035,457 "step": 56,458 "video_reward_cumulative_accuracy": 0.6339285714285714459 },460 {461 "epoch": 0.016918967052537846,462 "grad_norm": 5.258223056793213,463 "learning_rate": 8.456973293768548e-07,464 "loss": 0.0816,465 "step": 57,466 "video_reward_cumulative_accuracy": 0.6403508771929824467 },468 {469 "epoch": 0.017215791035915702,470 "grad_norm": 2.0548617839813232,471 "learning_rate": 8.605341246290802e-07,472 "loss": 0.083,473 "step": 58,474 "video_reward_cumulative_accuracy": 0.6379310344827587475 },476 {477 "epoch": 0.01751261501929356,478 "grad_norm": 6.927907466888428,479 "learning_rate": 8.753709198813057e-07,480 "loss": 0.1002,481 "step": 59,482 "video_reward_cumulative_accuracy": 0.635593220338983483 },484 {485 "epoch": 0.017809439002671415,486 "grad_norm": 6.766244411468506,487 "learning_rate": 8.902077151335312e-07,488 "loss": 0.096,489 "step": 60,490 "video_reward_cumulative_accuracy": 0.6416666666666667491 },492 {493 "epoch": 0.01810626298604927,494 "grad_norm": 3.500997304916382,495 "learning_rate": 9.050445103857568e-07,496 "loss": 0.0837,497 "step": 61,498 "video_reward_cumulative_accuracy": 0.6475409836065574499 },500 {501 "epoch": 0.01840308696942713,502 "grad_norm": 2.852778196334839,503 "learning_rate": 9.198813056379823e-07,504 "loss": 0.0913,505 "step": 62,506 "video_reward_cumulative_accuracy": 0.6451612903225806507 },508 {509 "epoch": 0.018699910952804988,510 "grad_norm": 10.269067764282227,511 "learning_rate": 9.347181008902077e-07,512 "loss": 0.0754,513 "step": 63,514 "video_reward_cumulative_accuracy": 0.6507936507936508515 },516 {517 "epoch": 0.018996734936182844,518 "grad_norm": 3.8665616512298584,519 "learning_rate": 9.495548961424333e-07,520 "loss": 0.0886,521 "step": 64,522 "video_reward_cumulative_accuracy": 0.65625523 },524 {525 "epoch": 0.0192935589195607,526 "grad_norm": 2.6777122020721436,527 "learning_rate": 9.643916913946588e-07,528 "loss": 0.0877,529 "step": 65,530 "video_reward_cumulative_accuracy": 0.6615384615384615531 },532 {533 "epoch": 0.019590382902938557,534 "grad_norm": 1.6913548707962036,535 "learning_rate": 9.792284866468842e-07,536 "loss": 0.0684,537 "step": 66,538 "video_reward_cumulative_accuracy": 0.6666666666666666539 },540 {541 "epoch": 0.019887206886316414,542 "grad_norm": 2.17547345161438,543 "learning_rate": 9.9406528189911e-07,544 "loss": 0.0903,545 "step": 67,546 "video_reward_cumulative_accuracy": 0.6716417910447762547 },548 {549 "epoch": 0.02018403086969427,550 "grad_norm": 9.660552978515625,551 "learning_rate": 1.0089020771513354e-06,552 "loss": 0.1137,553 "step": 68,554 "video_reward_cumulative_accuracy": 0.6617647058823529555 },556 {557 "epoch": 0.020480854853072127,558 "grad_norm": 6.081414222717285,559 "learning_rate": 1.0237388724035608e-06,560 "loss": 0.0836,561 "step": 69,562 "video_reward_cumulative_accuracy": 0.6594202898550725563 },564 {565 "epoch": 0.020777678836449986,566 "grad_norm": 5.362737655639648,567 "learning_rate": 1.0385756676557865e-06,568 "loss": 0.0828,569 "step": 70,570 "video_reward_cumulative_accuracy": 0.6642857142857143571 },572 {573 "epoch": 0.021074502819827843,574 "grad_norm": 3.4765052795410156,575 "learning_rate": 1.053412462908012e-06,576 "loss": 0.0678,577 "step": 71,578 "video_reward_cumulative_accuracy": 0.6619718309859155579 },580 {581 "epoch": 0.0213713268032057,582 "grad_norm": 3.868277072906494,583 "learning_rate": 1.0682492581602374e-06,584 "loss": 0.0704,585 "step": 72,586 "video_reward_cumulative_accuracy": 0.6666666666666666587 },588 {589 "epoch": 0.021668150786583556,590 "grad_norm": 2.056610107421875,591 "learning_rate": 1.083086053412463e-06,592 "loss": 0.0699,593 "step": 73,594 "video_reward_cumulative_accuracy": 0.6712328767123288595 },596 {597 "epoch": 0.021964974769961412,598 "grad_norm": 4.672041893005371,599 "learning_rate": 1.0979228486646885e-06,600 "loss": 0.0789,601 "step": 74,602 "video_reward_cumulative_accuracy": 0.668918918918919603 },604 {605 "epoch": 0.02226179875333927,606 "grad_norm": 10.368182182312012,607 "learning_rate": 1.112759643916914e-06,608 "loss": 0.0785,609 "step": 75,610 "video_reward_cumulative_accuracy": 0.6666666666666666611 },612 {613 "epoch": 0.022558622736717125,614 "grad_norm": 5.704588890075684,615 "learning_rate": 1.1275964391691395e-06,616 "loss": 0.0781,617 "step": 76,618 "video_reward_cumulative_accuracy": 0.6710526315789473619 },620 {621 "epoch": 0.022855446720094985,622 "grad_norm": 4.419633388519287,623 "learning_rate": 1.1424332344213651e-06,624 "loss": 0.0859,625 "step": 77,626 "video_reward_cumulative_accuracy": 0.6623376623376623627 },628 {629 "epoch": 0.02315227070347284,630 "grad_norm": 8.239113807678223,631 "learning_rate": 1.1572700296735906e-06,632 "loss": 0.0967,633 "step": 78,634 "video_reward_cumulative_accuracy": 0.6538461538461539635 },636 {637 "epoch": 0.023449094686850698,638 "grad_norm": 2.648416519165039,639 "learning_rate": 1.172106824925816e-06,640 "loss": 0.0906,641 "step": 79,642 "video_reward_cumulative_accuracy": 0.6582278481012658643 },644 {645 "epoch": 0.023745918670228554,646 "grad_norm": 6.028584003448486,647 "learning_rate": 1.1869436201780417e-06,648 "loss": 0.0789,649 "step": 80,650 "video_reward_cumulative_accuracy": 0.65625651 },652 {653 "epoch": 0.02404274265360641,654 "grad_norm": 4.417842864990234,655 "learning_rate": 1.2017804154302672e-06,656 "loss": 0.0756,657 "step": 81,658 "video_reward_cumulative_accuracy": 0.654320987654321659 },660 {661 "epoch": 0.024339566636984267,662 "grad_norm": 5.482295513153076,663 "learning_rate": 1.2166172106824927e-06,664 "loss": 0.0518,665 "step": 82,666 "video_reward_cumulative_accuracy": 0.6585365853658537667 },668 {669 "epoch": 0.024636390620362124,670 "grad_norm": 3.1841440200805664,671 "learning_rate": 1.2314540059347183e-06,672 "loss": 0.0881,673 "step": 83,674 "video_reward_cumulative_accuracy": 0.6566265060240963675 },676 {677 "epoch": 0.024933214603739984,678 "grad_norm": 5.477931976318359,679 "learning_rate": 1.2462908011869438e-06,680 "loss": 0.0712,681 "step": 84,682 "video_reward_cumulative_accuracy": 0.6547619047619048683 },684 {685 "epoch": 0.02523003858711784,686 "grad_norm": 2.8534862995147705,687 "learning_rate": 1.2611275964391693e-06,688 "loss": 0.0796,689 "step": 85,690 "video_reward_cumulative_accuracy": 0.6588235294117647691 },692 {693 "epoch": 0.025526862570495697,694 "grad_norm": 6.181415557861328,695 "learning_rate": 1.2759643916913947e-06,696 "loss": 0.0667,697 "step": 86,698 "video_reward_cumulative_accuracy": 0.6569767441860465699 },700 {701 "epoch": 0.025823686553873553,702 "grad_norm": 4.147750377655029,703 "learning_rate": 1.2908011869436202e-06,704 "loss": 0.0738,705 "step": 87,706 "video_reward_cumulative_accuracy": 0.6609195402298851707 },708 {709 "epoch": 0.02612051053725141,710 "grad_norm": 4.225668907165527,711 "learning_rate": 1.3056379821958458e-06,712 "loss": 0.1034,713 "step": 88,714 "video_reward_cumulative_accuracy": 0.6647727272727273715 },716 {717 "epoch": 0.026417334520629266,718 "grad_norm": 2.278589963912964,719 "learning_rate": 1.3204747774480713e-06,720 "loss": 0.0505,721 "step": 89,722 "video_reward_cumulative_accuracy": 0.6685393258426966723 },724 {725 "epoch": 0.026714158504007122,726 "grad_norm": 6.009743690490723,727 "learning_rate": 1.335311572700297e-06,728 "loss": 0.0665,729 "step": 90,730 "video_reward_cumulative_accuracy": 0.6722222222222223731 },732 {733 "epoch": 0.027010982487384982,734 "grad_norm": 3.622457265853882,735 "learning_rate": 1.3501483679525224e-06,736 "loss": 0.0803,737 "step": 91,738 "video_reward_cumulative_accuracy": 0.6703296703296703739 },740 {741 "epoch": 0.02730780647076284,742 "grad_norm": 2.3656132221221924,743 "learning_rate": 1.364985163204748e-06,744 "loss": 0.0679,745 "step": 92,746 "video_reward_cumulative_accuracy": 0.6739130434782609747 },748 {749 "epoch": 0.027604630454140695,750 "grad_norm": 5.818204879760742,751 "learning_rate": 1.3798219584569734e-06,752 "loss": 0.0931,753 "step": 93,754 "video_reward_cumulative_accuracy": 0.6720430107526881755 },756 {757 "epoch": 0.027901454437518552,758 "grad_norm": 3.5401792526245117,759 "learning_rate": 1.394658753709199e-06,760 "loss": 0.0723,761 "step": 94,762 "video_reward_cumulative_accuracy": 0.675531914893617763 },764 {765 "epoch": 0.028198278420896408,766 "grad_norm": 6.389036178588867,767 "learning_rate": 1.4094955489614245e-06,768 "loss": 0.0554,769 "step": 95,770 "video_reward_cumulative_accuracy": 0.6789473684210526771 },772 {773 "epoch": 0.028495102404274265,774 "grad_norm": 3.825948476791382,775 "learning_rate": 1.42433234421365e-06,776 "loss": 0.0705,777 "step": 96,778 "video_reward_cumulative_accuracy": 0.6822916666666666779 },780 {781 "epoch": 0.02879192638765212,782 "grad_norm": 3.565723180770874,783 "learning_rate": 1.4391691394658754e-06,784 "loss": 0.0737,785 "step": 97,786 "video_reward_cumulative_accuracy": 0.6855670103092784787 },788 {789 "epoch": 0.029088750371029978,790 "grad_norm": 2.599555253982544,791 "learning_rate": 1.4540059347181009e-06,792 "loss": 0.0468,793 "step": 98,794 "video_reward_cumulative_accuracy": 0.6887755102040817795 },796 {797 "epoch": 0.029385574354407838,798 "grad_norm": 2.7549595832824707,799 "learning_rate": 1.4688427299703265e-06,800 "loss": 0.0644,801 "step": 99,802 "video_reward_cumulative_accuracy": 0.6919191919191919803 },804 {805 "epoch": 0.029682398337785694,806 "grad_norm": 5.881991386413574,807 "learning_rate": 1.483679525222552e-06,808 "loss": 0.0955,809 "step": 100,810 "video_reward_cumulative_accuracy": 0.685811 },812 {813 "epoch": 0.02997922232116355,814 "grad_norm": 2.9013118743896484,815 "learning_rate": 1.4985163204747777e-06,816 "loss": 0.0603,817 "step": 101,818 "video_reward_cumulative_accuracy": 0.6881188118811881819 },820 {821 "epoch": 0.030276046304541407,822 "grad_norm": 3.3732762336730957,823 "learning_rate": 1.5133531157270031e-06,824 "loss": 0.0615,825 "step": 102,826 "video_reward_cumulative_accuracy": 0.6911764705882353827 },828 {829 "epoch": 0.030572870287919263,830 "grad_norm": 4.168172359466553,831 "learning_rate": 1.5281899109792286e-06,832 "loss": 0.1154,833 "step": 103,834 "video_reward_cumulative_accuracy": 0.6893203883495146835 },836 {837 "epoch": 0.03086969427129712,838 "grad_norm": 3.426560163497925,839 "learning_rate": 1.543026706231454e-06,840 "loss": 0.0481,841 "step": 104,842 "video_reward_cumulative_accuracy": 0.6923076923076923843 },844 {845 "epoch": 0.031166518254674976,846 "grad_norm": 8.7980375289917,847 "learning_rate": 1.5578635014836795e-06,848 "loss": 0.116,849 "step": 105,850 "video_reward_cumulative_accuracy": 0.6952380952380952851 },852 {853 "epoch": 0.03146334223805283,854 "grad_norm": 3.171637773513794,855 "learning_rate": 1.572700296735905e-06,856 "loss": 0.0592,857 "step": 106,858 "video_reward_cumulative_accuracy": 0.6933962264150944859 },860 {861 "epoch": 0.03176016622143069,862 "grad_norm": 2.1238746643066406,863 "learning_rate": 1.5875370919881309e-06,864 "loss": 0.0498,865 "step": 107,866 "video_reward_cumulative_accuracy": 0.6962616822429907867 },868 {869 "epoch": 0.032056990204808546,870 "grad_norm": 3.829890727996826,871 "learning_rate": 1.6023738872403563e-06,872 "loss": 0.0897,873 "step": 108,874 "video_reward_cumulative_accuracy": 0.6990740740740741875 },876 {877 "epoch": 0.032353814188186406,878 "grad_norm": 3.1230406761169434,879 "learning_rate": 1.6172106824925818e-06,880 "loss": 0.0446,881 "step": 109,882 "video_reward_cumulative_accuracy": 0.7018348623853211883 },884 {885 "epoch": 0.032650638171564265,886 "grad_norm": 5.97599983215332,887 "learning_rate": 1.6320474777448073e-06,888 "loss": 0.1308,889 "step": 110,890 "video_reward_cumulative_accuracy": 0.7045454545454546891 },892 {893 "epoch": 0.03294746215494212,894 "grad_norm": 2.300419330596924,895 "learning_rate": 1.6468842729970327e-06,896 "loss": 0.0632,897 "step": 111,898 "video_reward_cumulative_accuracy": 0.7072072072072072899 },900 {901 "epoch": 0.03324428613831998,902 "grad_norm": 2.3834316730499268,903 "learning_rate": 1.6617210682492582e-06,904 "loss": 0.0585,905 "step": 112,906 "video_reward_cumulative_accuracy": 0.7098214285714286907 },908 {909 "epoch": 0.03354111012169783,910 "grad_norm": 2.218229293823242,911 "learning_rate": 1.6765578635014836e-06,912 "loss": 0.0574,913 "step": 113,914 "video_reward_cumulative_accuracy": 0.7079646017699115915 },916 {917 "epoch": 0.03383793410507569,918 "grad_norm": 6.411158084869385,919 "learning_rate": 1.6913946587537095e-06,920 "loss": 0.0608,921 "step": 114,922 "video_reward_cumulative_accuracy": 0.706140350877193923 },924 {925 "epoch": 0.034134758088453544,926 "grad_norm": 3.1801657676696777,927 "learning_rate": 1.706231454005935e-06,928 "loss": 0.0894,929 "step": 115,930 "video_reward_cumulative_accuracy": 0.7043478260869566931 },932 {933 "epoch": 0.034431582071831404,934 "grad_norm": 6.5272536277771,935 "learning_rate": 1.7210682492581604e-06,936 "loss": 0.0919,937 "step": 116,938 "video_reward_cumulative_accuracy": 0.7025862068965517939 },940 {941 "epoch": 0.034728406055209264,942 "grad_norm": 4.742123603820801,943 "learning_rate": 1.735905044510386e-06,944 "loss": 0.0324,945 "step": 117,946 "video_reward_cumulative_accuracy": 0.7051282051282052947 },948 {949 "epoch": 0.03502523003858712,950 "grad_norm": 3.718932628631592,951 "learning_rate": 1.7507418397626114e-06,952 "loss": 0.0611,953 "step": 118,954 "video_reward_cumulative_accuracy": 0.7033898305084746955 },956 {957 "epoch": 0.03532205402196498,958 "grad_norm": 4.320330619812012,959 "learning_rate": 1.7655786350148368e-06,960 "loss": 0.054,961 "step": 119,962 "video_reward_cumulative_accuracy": 0.7058823529411765963 },964 {965 "epoch": 0.03561887800534283,966 "grad_norm": 4.672208786010742,967 "learning_rate": 1.7804154302670625e-06,968 "loss": 0.0525,969 "step": 120,970 "video_reward_cumulative_accuracy": 0.7083333333333334971 },972 {973 "epoch": 0.03591570198872069,974 "grad_norm": 2.6272125244140625,975 "learning_rate": 1.7952522255192882e-06,976 "loss": 0.0652,977 "step": 121,978 "video_reward_cumulative_accuracy": 0.7107438016528925979 },980 {981 "epoch": 0.03621252597209854,982 "grad_norm": 1.9446464776992798,983 "learning_rate": 1.8100890207715136e-06,984 "loss": 0.043,985 "step": 122,986 "video_reward_cumulative_accuracy": 0.7131147540983607987 },988 {989 "epoch": 0.0365093499554764,990 "grad_norm": 4.1938910484313965,991 "learning_rate": 1.824925816023739e-06,992 "loss": 0.1023,993 "step": 123,994 "video_reward_cumulative_accuracy": 0.7154471544715447995 },996 {997 "epoch": 0.03680617393885426,998 "grad_norm": 3.999626874923706,999 "learning_rate": 1.8397626112759646e-06,1000 "loss": 0.062,1001 "step": 124,1002 "video_reward_cumulative_accuracy": 0.71370967741935491003 },1004 {1005 "epoch": 0.037102997922232116,1006 "grad_norm": 2.076876640319824,1007 "learning_rate": 1.85459940652819e-06,1008 "loss": 0.0479,1009 "step": 125,1010 "video_reward_cumulative_accuracy": 0.7161011 },1012 {1013 "epoch": 0.037399821905609976,1014 "grad_norm": 6.920149326324463,1015 "learning_rate": 1.8694362017804155e-06,1016 "loss": 0.0538,1017 "step": 126,1018 "video_reward_cumulative_accuracy": 0.71825396825396831019 },1020 {1021 "epoch": 0.03769664588898783,1022 "grad_norm": 2.615006923675537,1023 "learning_rate": 1.8842729970326411e-06,1024 "loss": 0.0352,1025 "step": 127,1026 "video_reward_cumulative_accuracy": 0.72047244094488191027 },1028 {1029 "epoch": 0.03799346987236569,1030 "grad_norm": 2.167612075805664,1031 "learning_rate": 1.8991097922848666e-06,1032 "loss": 0.0397,1033 "step": 128,1034 "video_reward_cumulative_accuracy": 0.722656251035 },1036 {1037 "epoch": 0.03829029385574354,1038 "grad_norm": 3.5347766876220703,1039 "learning_rate": 1.9139465875370923e-06,1040 "loss": 0.0699,1041 "step": 129,1042 "video_reward_cumulative_accuracy": 0.72093023255813951043 },1044 {1045 "epoch": 0.0385871178391214,1046 "grad_norm": 6.354689121246338,1047 "learning_rate": 1.9287833827893175e-06,1048 "loss": 0.0554,1049 "step": 130,1050 "video_reward_cumulative_accuracy": 0.71923076923076921051 },1052 {1053 "epoch": 0.03888394182249926,1054 "grad_norm": 2.4313793182373047,1055 "learning_rate": 1.943620178041543e-06,1056 "loss": 0.0529,1057 "step": 131,1058 "video_reward_cumulative_accuracy": 0.71755725190839691059 },1060 {1061 "epoch": 0.039180765805877114,1062 "grad_norm": 5.276021480560303,1063 "learning_rate": 1.9584569732937684e-06,1064 "loss": 0.0958,1065 "step": 132,1066 "video_reward_cumulative_accuracy": 0.71590909090909091067 },1068 {1069 "epoch": 0.039477589789254974,1070 "grad_norm": 6.521552562713623,1071 "learning_rate": 1.9732937685459945e-06,1072 "loss": 0.0698,1073 "step": 133,1074 "video_reward_cumulative_accuracy": 0.71428571428571431075 },1076 {1077 "epoch": 0.03977441377263283,1078 "grad_norm": 4.296199798583984,1079 "learning_rate": 1.98813056379822e-06,1080 "loss": 0.0632,1081 "step": 134,1082 "video_reward_cumulative_accuracy": 0.70895522388059711083 },1084 {1085 "epoch": 0.04007123775601069,1086 "grad_norm": 5.775026321411133,1087 "learning_rate": 2.0029673590504455e-06,1088 "loss": 0.1013,1089 "step": 135,1090 "video_reward_cumulative_accuracy": 0.71111111111111111091 },1092 {1093 "epoch": 0.04036806173938854,1094 "grad_norm": 2.675187349319458,1095 "learning_rate": 2.0178041543026707e-06,1096 "loss": 0.0515,1097 "step": 136,1098 "video_reward_cumulative_accuracy": 0.71323529411764711099 },1100 {1101 "epoch": 0.0406648857227664,1102 "grad_norm": 6.260149955749512,1103 "learning_rate": 2.0326409495548964e-06,1104 "loss": 0.0883,1105 "step": 137,1106 "video_reward_cumulative_accuracy": 0.71532846715328471107 },1108 {1109 "epoch": 0.04096170970614425,1110 "grad_norm": 11.186830520629883,1111 "learning_rate": 2.0474777448071216e-06,1112 "loss": 0.0918,1113 "step": 138,1114 "video_reward_cumulative_accuracy": 0.7173913043478261115 },1116 {1117 "epoch": 0.04125853368952211,1118 "grad_norm": 6.043707847595215,1119 "learning_rate": 2.0623145400593473e-06,1120 "loss": 0.085,1121 "step": 139,1122 "video_reward_cumulative_accuracy": 0.71582733812949641123 },1124 {1125 "epoch": 0.04155535767289997,1126 "grad_norm": 9.900096893310547,1127 "learning_rate": 2.077151335311573e-06,1128 "loss": 0.0977,1129 "step": 140,1130 "video_reward_cumulative_accuracy": 0.71785714285714291131 },1132 {1133 "epoch": 0.041852181656277826,1134 "grad_norm": 2.8422248363494873,1135 "learning_rate": 2.0919881305637987e-06,1136 "loss": 0.0399,1137 "step": 141,1138 "video_reward_cumulative_accuracy": 0.71631205673758871139 },1140 {1141 "epoch": 0.042149005639655686,1142 "grad_norm": 2.6769790649414062,1143 "learning_rate": 2.106824925816024e-06,1144 "loss": 0.0355,1145 "step": 142,1146 "video_reward_cumulative_accuracy": 0.71830985915492961147 },1148 {1149 "epoch": 0.04244582962303354,1150 "grad_norm": 3.1609225273132324,1151 "learning_rate": 2.1216617210682496e-06,1152 "loss": 0.0677,1153 "step": 143,1154 "video_reward_cumulative_accuracy": 0.71328671328671331155 },1156 {1157 "epoch": 0.0427426536064114,1158 "grad_norm": 2.4717013835906982,1159 "learning_rate": 2.136498516320475e-06,1160 "loss": 0.0365,1161 "step": 144,1162 "video_reward_cumulative_accuracy": 0.71180555555555561163 },1164 {1165 "epoch": 0.04303947758978925,1166 "grad_norm": 8.373668670654297,1167 "learning_rate": 2.1513353115727005e-06,1168 "loss": 0.1185,1169 "step": 145,1170 "video_reward_cumulative_accuracy": 0.71034482758620691171 },1172 {1173 "epoch": 0.04333630157316711,1174 "grad_norm": 2.4484148025512695,1175 "learning_rate": 2.166172106824926e-06,1176 "loss": 0.0479,1177 "step": 146,1178 "video_reward_cumulative_accuracy": 0.71232876712328761179 },1180 {1181 "epoch": 0.04363312555654497,1182 "grad_norm": 4.690200328826904,1183 "learning_rate": 2.1810089020771514e-06,1184 "loss": 0.0706,1185 "step": 147,1186 "video_reward_cumulative_accuracy": 0.71428571428571431187 },1188 {1189 "epoch": 0.043929949539922825,1190 "grad_norm": 5.8376898765563965,1191 "learning_rate": 2.195845697329377e-06,1192 "loss": 0.0849,1193 "step": 148,1194 "video_reward_cumulative_accuracy": 0.71621621621621621195 },1196 {1197 "epoch": 0.044226773523300685,1198 "grad_norm": 2.741074562072754,1199 "learning_rate": 2.2106824925816028e-06,1200 "loss": 0.0741,