smolagents/SmolVLM2-2.2B-Instruct-Agentic-GUI
67232
1{2 "best_global_step": null,3 "best_metric": null,4 "best_model_checkpoint": null,5 "epoch": 2.0,6 "eval_steps": 100,7 "global_step": 1408,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "epoch": 0.007109215320359016,14 "grad_norm": 22.657577985866105,15 "learning_rate": 9.302325581395349e-06,16 "loss": 2.574,17 "mean_token_accuracy": 0.5464246176183224,18 "num_tokens": 4589382.0,19 "step": 520 },21 {22 "epoch": 0.014218430640718031,23 "grad_norm": 2.3543289370923013,24 "learning_rate": 2.0930232558139536e-05,25 "loss": 1.4882,26 "mean_token_accuracy": 0.6589333653450012,27 "num_tokens": 9171524.0,28 "step": 1029 },30 {31 "epoch": 0.021327645961077047,32 "grad_norm": 0.8063547574982903,33 "learning_rate": 3.2558139534883724e-05,34 "loss": 1.0174,35 "mean_token_accuracy": 0.7330243036150932,36 "num_tokens": 13765157.0,37 "step": 1538 },39 {40 "epoch": 0.028436861281436063,41 "grad_norm": 0.572573905518242,42 "learning_rate": 4.418604651162791e-05,43 "loss": 0.8773,44 "mean_token_accuracy": 0.7569610200822353,45 "num_tokens": 18369874.0,46 "step": 2047 },48 {49 "epoch": 0.035546076601795075,50 "grad_norm": 0.5738482260117446,51 "learning_rate": 5.5813953488372095e-05,52 "loss": 0.7975,53 "mean_token_accuracy": 0.7729738861322403,54 "num_tokens": 22960290.0,55 "step": 2556 },57 {58 "epoch": 0.042655291922154094,59 "grad_norm": 0.5016568944917689,60 "learning_rate": 6.744186046511628e-05,61 "loss": 0.7632,62 "mean_token_accuracy": 0.778630904853344,63 "num_tokens": 27556623.0,64 "step": 3065 },66 {67 "epoch": 0.049764507242513106,68 "grad_norm": 0.4845613474361907,69 "learning_rate": 7.906976744186047e-05,70 "loss": 0.7326,71 "mean_token_accuracy": 0.7872321248054505,72 "num_tokens": 32158408.0,73 "step": 3574 },75 {76 "epoch": 0.056873722562872125,77 "grad_norm": 0.4270154516127363,78 "learning_rate": 9.069767441860465e-05,79 "loss": 0.7095,80 "mean_token_accuracy": 0.7919960044324398,81 "num_tokens": 36742233.0,82 "step": 4083 },84 {85 "epoch": 0.06398293788323114,86 "grad_norm": 0.499498695141066,87 "learning_rate": 9.9999880816326e-05,88 "loss": 0.6973,89 "mean_token_accuracy": 0.7952379912137986,90 "num_tokens": 41335670.0,91 "step": 4592 },93 {94 "epoch": 0.07109215320359015,95 "grad_norm": 0.4645180201543763,96 "learning_rate": 9.999570945402425e-05,97 "loss": 0.6853,98 "mean_token_accuracy": 0.7981184311211109,99 "num_tokens": 45940079.0,100 "step": 50101 },102 {103 "epoch": 0.07820136852394917,104 "grad_norm": 0.434255531179794,105 "learning_rate": 9.998557953932929e-05,106 "loss": 0.6688,107 "mean_token_accuracy": 0.8012012615799904,108 "num_tokens": 50533771.0,109 "step": 55110 },111 {112 "epoch": 0.08531058384430819,113 "grad_norm": 0.393754634337621,114 "learning_rate": 9.99694924136941e-05,115 "loss": 0.6725,116 "mean_token_accuracy": 0.800255061686039,117 "num_tokens": 55133444.0,118 "step": 60119 },120 {121 "epoch": 0.0924197991646672,122 "grad_norm": 0.49718727212066355,123 "learning_rate": 9.99474502074547e-05,124 "loss": 0.6664,125 "mean_token_accuracy": 0.801218880712986,126 "num_tokens": 59726447.0,127 "step": 65128 },129 {130 "epoch": 0.09952901448502621,131 "grad_norm": 0.4005142024066312,132 "learning_rate": 9.991945583954808e-05,133 "loss": 0.6549,134 "mean_token_accuracy": 0.8056452445685863,135 "num_tokens": 64319917.0,136 "step": 70137 },138 {139 "epoch": 0.10663822980538523,140 "grad_norm": 0.3774090383980249,141 "learning_rate": 9.988551301712567e-05,142 "loss": 0.6454,143 "mean_token_accuracy": 0.806719920784235,144 "num_tokens": 68898868.0,145 "step": 75146 },147 {148 "epoch": 0.11374744512574425,149 "grad_norm": 0.3995895890256704,150 "learning_rate": 9.984562623506235e-05,151 "loss": 0.6464,152 "mean_token_accuracy": 0.8064703330397606,153 "num_tokens": 73481972.0,154 "step": 80155 },156 {157 "epoch": 0.12085666044610326,158 "grad_norm": 0.3801619159341505,159 "learning_rate": 9.979980077536136e-05,160 "loss": 0.6462,161 "mean_token_accuracy": 0.8080633491277694,162 "num_tokens": 78079419.0,163 "step": 85164 },165 {166 "epoch": 0.1279658757664623,167 "grad_norm": 0.37074794226689833,168 "learning_rate": 9.974804270645462e-05,169 "loss": 0.6362,170 "mean_token_accuracy": 0.8091117829084397,171 "num_tokens": 82670195.0,172 "step": 90173 },174 {175 "epoch": 0.13507509108682128,176 "grad_norm": 0.37193721608812236,177 "learning_rate": 9.969035888239937e-05,178 "loss": 0.635,179 "mean_token_accuracy": 0.8079991653561592,180 "num_tokens": 87257953.0,181 "step": 95182 },183 {184 "epoch": 0.1421843064071803,185 "grad_norm": 0.36251703620037773,186 "learning_rate": 9.96267569419703e-05,187 "loss": 0.6315,188 "mean_token_accuracy": 0.8096475720405578,189 "num_tokens": 91838382.0,190 "step": 100191 },192 {193 "epoch": 0.1421843064071803,194 "eval_loss": 0.5971412062644958,195 "eval_mean_token_accuracy": 0.8093206621052926,196 "eval_num_tokens": 91838382.0,197 "eval_runtime": 141.8153,198 "eval_samples_per_second": 25.653,199 "eval_steps_per_second": 0.804,200 "step": 100201 },202 {203 "epoch": 0.14929352172753932,204 "grad_norm": 0.41583625971563776,205 "learning_rate": 9.955724530764809e-05,206 "loss": 0.6381,207 "mean_token_accuracy": 0.8077230393886566,208 "num_tokens": 96431755.0,209 "step": 105210 },211 {212 "epoch": 0.15640273704789834,213 "grad_norm": 0.3705693803444073,214 "learning_rate": 9.948183318450413e-05,215 "loss": 0.6197,216 "mean_token_accuracy": 0.8116156131029129,217 "num_tokens": 101027690.0,218 "step": 110219 },220 {221 "epoch": 0.16351195236825736,222 "grad_norm": 0.3214510651452395,223 "learning_rate": 9.940053055898133e-05,224 "loss": 0.6313,225 "mean_token_accuracy": 0.8089181430637836,226 "num_tokens": 105628547.0,227 "step": 115228 },229 {230 "epoch": 0.17062116768861638,231 "grad_norm": 0.34220731720085373,232 "learning_rate": 9.93133481975719e-05,233 "loss": 0.6077,234 "mean_token_accuracy": 0.814984206855297,235 "num_tokens": 110243592.0,236 "step": 120237 },238 {239 "epoch": 0.1777303830089754,240 "grad_norm": 0.35675802487560043,241 "learning_rate": 9.922029764539148e-05,242 "loss": 0.6263,243 "mean_token_accuracy": 0.8096928559243679,244 "num_tokens": 114832845.0,245 "step": 125246 },247 {248 "epoch": 0.1848395983293344,249 "grad_norm": 0.3422296936678833,250 "learning_rate": 9.912139122465027e-05,251 "loss": 0.6116,252 "mean_token_accuracy": 0.8140982151031494,253 "num_tokens": 119435421.0,254 "step": 130255 },256 {257 "epoch": 0.1919488136496934,258 "grad_norm": 0.3599918244922273,259 "learning_rate": 9.901664203302126e-05,260 "loss": 0.6052,261 "mean_token_accuracy": 0.8154805108904839,262 "num_tokens": 124028647.0,263 "step": 135264 },265 {266 "epoch": 0.19905802897005243,267 "grad_norm": 0.3595154303423279,268 "learning_rate": 9.890606394190588e-05,269 "loss": 0.6126,270 "mean_token_accuracy": 0.8132404424250126,271 "num_tokens": 128628413.0,272 "step": 140273 },274 {275 "epoch": 0.20616724429041144,276 "grad_norm": 0.3711012466200944,277 "learning_rate": 9.878967159459693e-05,278 "loss": 0.6068,279 "mean_token_accuracy": 0.8164977565407753,280 "num_tokens": 133219422.0,281 "step": 145282 },283 {284 "epoch": 0.21327645961077046,285 "grad_norm": 0.35910926284617867,286 "learning_rate": 9.866748040433956e-05,287 "loss": 0.6099,288 "mean_token_accuracy": 0.8152773261070252,289 "num_tokens": 137825952.0,290 "step": 150291 },292 {293 "epoch": 0.22038567493112948,294 "grad_norm": 0.4205439208166243,295 "learning_rate": 9.853950655229009e-05,296 "loss": 0.6064,297 "mean_token_accuracy": 0.815191026777029,298 "num_tokens": 142422368.0,299 "step": 155300 },301 {302 "epoch": 0.2274948902514885,303 "grad_norm": 0.32091150374802263,304 "learning_rate": 9.840576698537329e-05,305 "loss": 0.6093,306 "mean_token_accuracy": 0.8135301224887371,307 "num_tokens": 147015990.0,308 "step": 160309 },310 {311 "epoch": 0.23460410557184752,312 "grad_norm": 0.32627028158119226,313 "learning_rate": 9.826627941403811e-05,314 "loss": 0.5969,315 "mean_token_accuracy": 0.8182829037308693,316 "num_tokens": 151627096.0,317 "step": 165318 },319 {320 "epoch": 0.2417133208922065,321 "grad_norm": 0.32405674248273,322 "learning_rate": 9.812106230991248e-05,323 "loss": 0.6068,324 "mean_token_accuracy": 0.8159149341285229,325 "num_tokens": 156218968.0,326 "step": 170327 },328 {329 "epoch": 0.24882253621256553,330 "grad_norm": 0.3206982540127891,331 "learning_rate": 9.79701349033571e-05,332 "loss": 0.6039,333 "mean_token_accuracy": 0.8161494679749012,334 "num_tokens": 160797401.0,335 "step": 175336 },337 {338 "epoch": 0.2559317515329246,339 "grad_norm": 0.3360732448004463,340 "learning_rate": 9.78135171809189e-05,341 "loss": 0.6068,342 "mean_token_accuracy": 0.8159954428672791,343 "num_tokens": 165402684.0,344 "step": 180345 },346 {347 "epoch": 0.26304096685328354,348 "grad_norm": 0.33789233259366286,349 "learning_rate": 9.76512298826844e-05,350 "loss": 0.6026,351 "mean_token_accuracy": 0.8167447924613953,352 "num_tokens": 169997282.0,353 "step": 185354 },355 {356 "epoch": 0.27015018217364256,357 "grad_norm": 0.3089560668153988,358 "learning_rate": 9.748329449953302e-05,359 "loss": 0.5904,360 "mean_token_accuracy": 0.8193566597998142,361 "num_tokens": 174589836.0,362 "step": 190363 },364 {365 "epoch": 0.2772593974940016,366 "grad_norm": 0.32060053414915524,367 "learning_rate": 9.73097332702914e-05,368 "loss": 0.6044,369 "mean_token_accuracy": 0.8175870932638645,370 "num_tokens": 179181747.0,371 "step": 195372 },373 {374 "epoch": 0.2843686128143606,375 "grad_norm": 0.32004664048912745,376 "learning_rate": 9.713056917878818e-05,377 "loss": 0.5888,378 "mean_token_accuracy": 0.8192018747329712,379 "num_tokens": 183760367.0,380 "step": 200381 },382 {383 "epoch": 0.2843686128143606,384 "eval_loss": 0.5599971413612366,385 "eval_mean_token_accuracy": 0.8188303473748659,386 "eval_num_tokens": 183760367.0,387 "eval_runtime": 145.8536,388 "eval_samples_per_second": 24.943,389 "eval_steps_per_second": 0.782,390 "step": 200391 },392 {393 "epoch": 0.2914778281347196,394 "grad_norm": 0.3094551492752116,395 "learning_rate": 9.694582595081057e-05,396 "loss": 0.5872,397 "mean_token_accuracy": 0.819921114295721,398 "num_tokens": 188360903.0,399 "step": 205400 },401 {402 "epoch": 0.29858704345507864,403 "grad_norm": 0.36254147904822126,404 "learning_rate": 9.67555280509623e-05,405 "loss": 0.5942,406 "mean_token_accuracy": 0.817745155096054,407 "num_tokens": 192932381.0,408 "step": 210409 },410 {411 "epoch": 0.30569625877543766,412 "grad_norm": 0.3377909564779145,413 "learning_rate": 9.655970067942405e-05,414 "loss": 0.5994,415 "mean_token_accuracy": 0.8163805276155471,416 "num_tokens": 197505985.0,417 "step": 215418 },419 {420 "epoch": 0.3128054740957967,421 "grad_norm": 0.30751780494672465,422 "learning_rate": 9.63583697686162e-05,423 "loss": 0.5902,424 "mean_token_accuracy": 0.8196643941104412,425 "num_tokens": 202105424.0,426 "step": 220427 },428 {429 "epoch": 0.3199146894161557,430 "grad_norm": 0.34345028355301316,431 "learning_rate": 9.615156197976477e-05,432 "loss": 0.582,433 "mean_token_accuracy": 0.8217154465615749,434 "num_tokens": 206686951.0,435 "step": 225436 },437 {438 "epoch": 0.3270239047365147,439 "grad_norm": 0.3216135018716631,440 "learning_rate": 9.593930469937087e-05,441 "loss": 0.5708,442 "mean_token_accuracy": 0.8250658005475998,443 "num_tokens": 211278788.0,444 "step": 230445 },446 {447 "epoch": 0.33413312005687373,448 "grad_norm": 0.32564659909940696,449 "learning_rate": 9.572162603558393e-05,450 "loss": 0.5928,451 "mean_token_accuracy": 0.819525595754385,452 "num_tokens": 215877205.0,453 "step": 235454 },455 {456 "epoch": 0.34124233537723275,457 "grad_norm": 0.4839583335140069,458 "learning_rate": 9.549855481447954e-05,459 "loss": 0.5882,460 "mean_token_accuracy": 0.8204580388963223,461 "num_tokens": 220486454.0,462 "step": 240463 },464 {465 "epoch": 0.34835155069759177,466 "grad_norm": 0.3268671171699921,467 "learning_rate": 9.527012057624224e-05,468 "loss": 0.5836,469 "mean_token_accuracy": 0.8208626843988895,470 "num_tokens": 225080225.0,471 "step": 245472 },473 {474 "epoch": 0.3554607660179508,475 "grad_norm": 0.3244498327733708,476 "learning_rate": 9.50363535712535e-05,477 "loss": 0.586,478 "mean_token_accuracy": 0.8207595020532608,479 "num_tokens": 229657012.0,480 "step": 250481 },482 {483 "epoch": 0.3625699813383098,484 "grad_norm": 0.29889265357291406,485 "learning_rate": 9.479728475608593e-05,486 "loss": 0.5919,487 "mean_token_accuracy": 0.8190862230956555,488 "num_tokens": 234248976.0,489 "step": 255490 },491 {492 "epoch": 0.3696791966586688,493 "grad_norm": 0.34636883393423384,494 "learning_rate": 9.455294578940384e-05,495 "loss": 0.5765,496 "mean_token_accuracy": 0.8226364821195602,497 "num_tokens": 238829734.0,498 "step": 260499 },500 {501 "epoch": 0.3767884119790278,502 "grad_norm": 0.3092592234408446,503 "learning_rate": 9.430336902777083e-05,504 "loss": 0.576,505 "mean_token_accuracy": 0.821333235502243,506 "num_tokens": 243418989.0,507 "step": 265508 },509 {510 "epoch": 0.3838976272993868,511 "grad_norm": 0.30454136223380207,512 "learning_rate": 9.404858752136499e-05,513 "loss": 0.5771,514 "mean_token_accuracy": 0.8237294301390647,515 "num_tokens": 248015701.0,516 "step": 270517 },518 {519 "epoch": 0.39100684261974583,520 "grad_norm": 0.30289215095264577,521 "learning_rate": 9.378863500960222e-05,522 "loss": 0.5709,523 "mean_token_accuracy": 0.8236084163188935,524 "num_tokens": 252613191.0,525 "step": 275526 },527 {528 "epoch": 0.39811605794010485,529 "grad_norm": 0.3010273864601919,530 "learning_rate": 9.352354591666827e-05,531 "loss": 0.5861,532 "mean_token_accuracy": 0.820894256979227,533 "num_tokens": 257210808.0,534 "step": 280535 },536 {537 "epoch": 0.40522527326046387,538 "grad_norm": 0.30175911100812025,539 "learning_rate": 9.325335534696017e-05,540 "loss": 0.5753,541 "mean_token_accuracy": 0.8225005254149437,542 "num_tokens": 261790131.0,543 "step": 285544 },545 {546 "epoch": 0.4123344885808229,547 "grad_norm": 0.28871941798325856,548 "learning_rate": 9.29780990804375e-05,549 "loss": 0.5799,550 "mean_token_accuracy": 0.821347926557064,551 "num_tokens": 266377324.0,552 "step": 290553 },554 {555 "epoch": 0.4194437039011819,556 "grad_norm": 0.28095014086273895,557 "learning_rate": 9.269781356788424e-05,558 "loss": 0.581,559 "mean_token_accuracy": 0.8209108576178551,560 "num_tokens": 270967910.0,561 "step": 295562 },563 {564 "epoch": 0.4265529192215409,565 "grad_norm": 0.2893211807696515,566 "learning_rate": 9.241253592608183e-05,567 "loss": 0.5755,568 "mean_token_accuracy": 0.8242007777094841,569 "num_tokens": 275570273.0,570 "step": 300571 },572 {573 "epoch": 0.4265529192215409,574 "eval_loss": 0.5416839122772217,575 "eval_mean_token_accuracy": 0.8231211885025627,576 "eval_num_tokens": 275570273.0,577 "eval_runtime": 145.5254,578 "eval_samples_per_second": 24.999,579 "eval_steps_per_second": 0.783,580 "step": 300581 },582 {583 "epoch": 0.43366213454189995,584 "grad_norm": 0.30733885282429685,585 "learning_rate": 9.212230393289385e-05,586 "loss": 0.5781,587 "mean_token_accuracy": 0.8230207331478596,588 "num_tokens": 280172533.0,589 "step": 305590 },591 {592 "epoch": 0.44077134986225897,593 "grad_norm": 0.2682470819307261,594 "learning_rate": 9.182715602226341e-05,595 "loss": 0.5625,596 "mean_token_accuracy": 0.8270745746791363,597 "num_tokens": 284763929.0,598 "step": 310599 },600 {601 "epoch": 0.447880565182618,602 "grad_norm": 0.2962012849994535,603 "learning_rate": 9.152713127912355e-05,604 "loss": 0.5848,605 "mean_token_accuracy": 0.8201167277991772,606 "num_tokens": 289376903.0,607 "step": 315608 },609 {610 "epoch": 0.454989780502977,611 "grad_norm": 0.28564514411407316,612 "learning_rate": 9.12222694342213e-05,613 "loss": 0.5732,614 "mean_token_accuracy": 0.8246621482074261,615 "num_tokens": 293966796.0,616 "step": 320617 },618 {619 "epoch": 0.462098995823336,620 "grad_norm": 0.30020425973519915,621 "learning_rate": 9.091261085885646e-05,622 "loss": 0.5606,623 "mean_token_accuracy": 0.826822079718113,624 "num_tokens": 298540346.0,625 "step": 325626 },627 {628 "epoch": 0.46920821114369504,629 "grad_norm": 0.2887047887642146,630 "learning_rate": 9.059819655953536e-05,631 "loss": 0.5738,632 "mean_token_accuracy": 0.823461939394474,633 "num_tokens": 303112604.0,634 "step": 330635 },636 {637 "epoch": 0.476317426464054,638 "grad_norm": 0.3180269352697689,639 "learning_rate": 9.027906817254063e-05,640 "loss": 0.5654,641 "mean_token_accuracy": 0.8256018176674843,642 "num_tokens": 307694241.0,643 "step": 335644 },645 {646 "epoch": 0.483426641784413,647 "grad_norm": 0.29567931374872014,648 "learning_rate": 8.995526795841753e-05,649 "loss": 0.558,650 "mean_token_accuracy": 0.8256605207920075,651 "num_tokens": 312289299.0,652 "step": 340653 },654 {655 "epoch": 0.49053585710477204,656 "grad_norm": 0.3336504103662035,657 "learning_rate": 8.962683879637747e-05,658 "loss": 0.5617,659 "mean_token_accuracy": 0.8257805988192558,660 "num_tokens": 316884766.0,661 "step": 345662 },663 {664 "epoch": 0.49764507242513106,665 "grad_norm": 0.3705167375534613,666 "learning_rate": 8.929382417861991e-05,667 "loss": 0.561,668 "mean_token_accuracy": 0.8267210200428963,669 "num_tokens": 321461198.0,670 "step": 350671 },672 {673 "epoch": 0.5047542877454901,674 "grad_norm": 0.2946584460529412,675 "learning_rate": 8.895626820457283e-05,676 "loss": 0.557,677 "mean_token_accuracy": 0.828194110840559,678 "num_tokens": 326064722.0,679 "step": 355680 },681 {682 "epoch": 0.5118635030658492,683 "grad_norm": 0.31227448766803945,684 "learning_rate": 8.861421557505282e-05,685 "loss": 0.5522,686 "mean_token_accuracy": 0.8295037761330605,687 "num_tokens": 330652094.0,688 "step": 360689 },690 {691 "epoch": 0.5189727183862082,692 "grad_norm": 1.0759474066945163,693 "learning_rate": 8.826771158634567e-05,694 "loss": 0.5629,695 "mean_token_accuracy": 0.8260238766670227,696 "num_tokens": 335255835.0,697 "step": 365698 },699 {700 "epoch": 0.5260819337065671,701 "grad_norm": 0.2758992633553522,702 "learning_rate": 8.791680212420797e-05,703 "loss": 0.5502,704 "mean_token_accuracy": 0.828965923935175,705 "num_tokens": 339843476.0,706 "step": 370707 },708 {709 "epoch": 0.5331911490269261,710 "grad_norm": 0.29696149610793166,711 "learning_rate": 8.756153365779066e-05,712 "loss": 0.5542,713 "mean_token_accuracy": 0.8278730027377605,714 "num_tokens": 344420533.0,715 "step": 375716 },717 {718 "epoch": 0.5403003643472851,719 "grad_norm": 0.284706804181623,720 "learning_rate": 8.720195323348545e-05,721 "loss": 0.559,722 "mean_token_accuracy": 0.8278782211244107,723 "num_tokens": 349010370.0,724 "step": 380725 },726 {727 "epoch": 0.5474095796676441,728 "grad_norm": 0.3046957362601185,729 "learning_rate": 8.68381084686946e-05,730 "loss": 0.5576,731 "mean_token_accuracy": 0.8258513130247593,732 "num_tokens": 353598451.0,733 "step": 385734 },735 {736 "epoch": 0.5545187949880032,737 "grad_norm": 0.3134773718519533,738 "learning_rate": 8.647004754552526e-05,739 "loss": 0.5612,740 "mean_token_accuracy": 0.8255665130913258,741 "num_tokens": 358195615.0,742 "step": 390743 },744 {745 "epoch": 0.5616280103083622,746 "grad_norm": 0.33349640254961,747 "learning_rate": 8.609781920440891e-05,748 "loss": 0.552,749 "mean_token_accuracy": 0.8278413727879524,750 "num_tokens": 362764034.0,751 "step": 395752 },753 {754 "epoch": 0.5687372256287212,755 "grad_norm": 0.32034152048464726,756 "learning_rate": 8.5721472737647e-05,757 "loss": 0.5534,758 "mean_token_accuracy": 0.8273369200527668,759 "num_tokens": 367350265.0,760 "step": 400761 },762 {763 "epoch": 0.5687372256287212,764 "eval_loss": 0.5274047255516052,765 "eval_mean_token_accuracy": 0.8264030280866121,766 "eval_num_tokens": 367350265.0,767 "eval_runtime": 146.0134,768 "eval_samples_per_second": 24.916,769 "eval_steps_per_second": 0.781,770 "step": 400771 },772 {773 "epoch": 0.5758464409490802,774 "grad_norm": 0.29085093151843905,775 "learning_rate": 8.534105798288331e-05,776 "loss": 0.5506,777 "mean_token_accuracy": 0.830031219124794,778 "num_tokens": 371939618.0,779 "step": 405780 },781 {782 "epoch": 0.5829556562694392,783 "grad_norm": 0.27710417408529203,784 "learning_rate": 8.49566253165043e-05,785 "loss": 0.5439,786 "mean_token_accuracy": 0.8304261237382888,787 "num_tokens": 376519800.0,788 "step": 410789 },790 {791 "epoch": 0.5900648715897983,792 "grad_norm": 0.2611394917691902,793 "learning_rate": 8.456822564696789e-05,794 "loss": 0.5409,795 "mean_token_accuracy": 0.832954341173172,796 "num_tokens": 381102299.0,797 "step": 415798 },799 {800 "epoch": 0.5971740869101573,801 "grad_norm": 0.42771473321829473,802 "learning_rate": 8.417591040806213e-05,803 "loss": 0.5504,804 "mean_token_accuracy": 0.8300940133631229,805 "num_tokens": 385700779.0,806 "step": 420807 },808 {809 "epoch": 0.6042833022305163,810 "grad_norm": 0.28194050483515865,811 "learning_rate": 8.377973155209387e-05,812 "loss": 0.5553,813 "mean_token_accuracy": 0.8270630918443203,814 "num_tokens": 390294365.0,815 "step": 425816 },817 {818 "epoch": 0.6113925175508753,819 "grad_norm": 0.27563889901609234,820 "learning_rate": 8.337974154300913e-05,821 "loss": 0.5427,822 "mean_token_accuracy": 0.8309814311563969,823 "num_tokens": 394889149.0,824 "step": 430825 },826 {827 "epoch": 0.6185017328712343,828 "grad_norm": 0.27875362292884753,829 "learning_rate": 8.297599334944542e-05,830 "loss": 0.5561,831 "mean_token_accuracy": 0.8275676898658275,832 "num_tokens": 399459807.0,833 "step": 435834 },835 {836 "epoch": 0.6256109481915934,837 "grad_norm": 0.7336148967265075,838 "learning_rate": 8.256854043771754e-05,839 "loss": 0.5507,840 "mean_token_accuracy": 0.8285100273787975,841 "num_tokens": 404034333.0,842 "step": 440843 },844 {845 "epoch": 0.6327201635119524,846 "grad_norm": 0.3259646654441019,847 "learning_rate": 8.215743676473719e-05,848 "loss": 0.5503,849 "mean_token_accuracy": 0.8290993146598339,850 "num_tokens": 408627270.0,851 "step": 445852 },853 {854 "epoch": 0.6398293788323114,855 "grad_norm": 0.3012299941832976,856 "learning_rate": 8.174273677086779e-05,857 "loss": 0.552,858 "mean_token_accuracy": 0.8279682919383049,859 "num_tokens": 413222911.0,860 "step": 450861 },862 {863 "epoch": 0.6469385941526704,864 "grad_norm": 0.30771992691522176,865 "learning_rate": 8.132449537271519e-05,866 "loss": 0.552,867 "mean_token_accuracy": 0.8296807646751404,868 "num_tokens": 417806274.0,869 "step": 455870 },871 {872 "epoch": 0.6540478094730294,873 "grad_norm": 0.2810763807856677,874 "learning_rate": 8.090276795585531e-05,875 "loss": 0.5414,876 "mean_token_accuracy": 0.8314659893512726,877 "num_tokens": 422401434.0,878 "step": 460879 },880 {881 "epoch": 0.6611570247933884,882 "grad_norm": 0.2672336811508722,883 "learning_rate": 8.047761036749985e-05,884 "loss": 0.5564,885 "mean_token_accuracy": 0.8265900291502476,886 "num_tokens": 426986385.0,887 "step": 465888 },889 {890 "epoch": 0.6682662401137475,891 "grad_norm": 0.25924906311163326,892 "learning_rate": 8.004907890910055e-05,893 "loss": 0.5452,894 "mean_token_accuracy": 0.8297064855694771,895 "num_tokens": 431585703.0,896 "step": 470897 },898 {899 "epoch": 0.6753754554341065,900 "grad_norm": 0.2772688573388134,901 "learning_rate": 7.961723032889358e-05,902 "loss": 0.5292,903 "mean_token_accuracy": 0.8346129797399044,904 "num_tokens": 436150194.0,905 "step": 475906 },907 {908 "epoch": 0.6824846707544655,909 "grad_norm": 0.25573353155086187,910 "learning_rate": 7.918212181438467e-05,911 "loss": 0.5397,912 "mean_token_accuracy": 0.8314497999846935,913 "num_tokens": 440736901.0,914 "step": 480915 },916 {917 "epoch": 0.6895938860748245,918 "grad_norm": 0.2640386419783165,919 "learning_rate": 7.874381098477599e-05,920 "loss": 0.5359,921 "mean_token_accuracy": 0.8328767582774163,922 "num_tokens": 445334774.0,923 "step": 485924 },925 {926 "epoch": 0.6967031013951835,927 "grad_norm": 0.2662269663206075,928 "learning_rate": 7.830235588333597e-05,929 "loss": 0.5578,930 "mean_token_accuracy": 0.8268053226172924,931 "num_tokens": 449908855.0,932 "step": 490933 },934 {935 "epoch": 0.7038123167155426,936 "grad_norm": 0.2756351015892551,937 "learning_rate": 7.785781496971297e-05,938 "loss": 0.5503,939 "mean_token_accuracy": 0.8284729138016701,940 "num_tokens": 454513487.0,941 "step": 495942 },943 {944 "epoch": 0.7109215320359016,945 "grad_norm": 0.4547105928976161,946 "learning_rate": 7.741024711219366e-05,947 "loss": 0.5431,948 "mean_token_accuracy": 0.8298681430518627,949 "num_tokens": 459106365.0,950 "step": 500951 },952 {953 "epoch": 0.7109215320359016,954 "eval_loss": 0.5168540477752686,955 "eval_mean_token_accuracy": 0.8290872861418808,956 "eval_num_tokens": 459106365.0,957 "eval_runtime": 146.2066,958 "eval_samples_per_second": 24.883,959 "eval_steps_per_second": 0.78,960 "step": 500961 },962 {963 "epoch": 0.7180307473562606,964 "grad_norm": 1.6021704699780053,965 "learning_rate": 7.695971157990754e-05,966 "loss": 0.5646,967 "mean_token_accuracy": 0.8263038910925389,968 "num_tokens": 463703240.0,969 "step": 505970 },971 {972 "epoch": 0.7251399626766196,973 "grad_norm": 4.625968090811763,974 "learning_rate": 7.650626803497806e-05,975 "loss": 0.5581,976 "mean_token_accuracy": 0.8270722553133965,977 "num_tokens": 468295660.0,978 "step": 510979 },980 {981 "epoch": 0.7322491779969785,982 "grad_norm": 0.27503115183353516,983 "learning_rate": 7.604997652462205e-05,984 "loss": 0.5492,985 "mean_token_accuracy": 0.8294327199459076,986 "num_tokens": 472896751.0,987 "step": 515988 },989 {990 "epoch": 0.7393583933173375,991 "grad_norm": 0.267416722991217,992 "learning_rate": 7.55908974731978e-05,993 "loss": 0.5418,994 "mean_token_accuracy": 0.8326966613531113,995 "num_tokens": 477480918.0,996 "step": 520997 },998 {999 "epoch": 0.7464676086376966,1000 "grad_norm": 0.25628361203172423,1001 "learning_rate": 7.512909167420347e-05,1002 "loss": 0.5404,1003 "mean_token_accuracy": 0.8324044570326805,1004 "num_tokens": 482064392.0,1005 "step": 5251006 },1007 {1008 "epoch": 0.7535768239580556,1009 "grad_norm": 0.24597696845219366,1010 "learning_rate": 7.466462028222654e-05,1011 "loss": 0.5353,1012 "mean_token_accuracy": 0.8331540204584599,1013 "num_tokens": 486649806.0,1014 "step": 5301015 },1016 {1017 "epoch": 0.7606860392784146,1018 "grad_norm": 0.2497969231256322,1019 "learning_rate": 7.419754480484536e-05,1020 "loss": 0.5378,1021 "mean_token_accuracy": 0.8323175966739654,1022 "num_tokens": 491217398.0,1023 "step": 5351024 },1025 {1026 "epoch": 0.7677952545987736,1027 "grad_norm": 0.27136426093422567,1028 "learning_rate": 7.3727927094484e-05,1029 "loss": 0.5303,1030 "mean_token_accuracy": 0.8346898458898068,1031 "num_tokens": 495798334.0,1032 "step": 5401033 },1034 {1035 "epoch": 0.7749044699191326,1036 "grad_norm": 0.263928683082665,1037 "learning_rate": 7.32558293402215e-05,1038 "loss": 0.5193,1039 "mean_token_accuracy": 0.8367893837392331,1040 "num_tokens": 500382331.0,1041 "step": 5451042 },1043 {1044 "epoch": 0.7820136852394917,1045 "grad_norm": 0.2697485453052082,1046 "learning_rate": 7.27813140595565e-05,1047 "loss": 0.5249,1048 "mean_token_accuracy": 0.836308328807354,1049 "num_tokens": 504972961.0,1050 "step": 5501051 },1052 {1053 "epoch": 0.7891229005598507,1054 "grad_norm": 0.47577994241811294,1055 "learning_rate": 7.23044440901283e-05,1056 "loss": 0.5386,1057 "mean_token_accuracy": 0.832004614919424,1058 "num_tokens": 509556175.0,1059 "step": 5551060 },1061 {1062 "epoch": 0.7962321158802097,1063 "grad_norm": 0.26812210950339255,1064 "learning_rate": 7.182528258139563e-05,1065 "loss": 0.5327,1066 "mean_token_accuracy": 0.8331871695816517,1067 "num_tokens": 514159170.0,1068 "step": 5601069 },1070 {1071 "epoch": 0.8033413312005687,1072 "grad_norm": 0.2590503131411491,1073 "learning_rate": 7.13438929862741e-05,1074 "loss": 0.5447,1075 "mean_token_accuracy": 0.8303000062704087,1076 "num_tokens": 518758083.0,1077 "step": 5651078 },1079 {1080 "epoch": 0.8104505465209277,1081 "grad_norm": 0.2700164600845211,1082 "learning_rate": 7.086033905273344e-05,1083 "loss": 0.5367,1084 "mean_token_accuracy": 0.8323484763503075,1085 "num_tokens": 523345629.0,1086 "step": 5701087 },1088 {1089 "epoch": 0.8175597618412868,1090 "grad_norm": 0.26967028018820877,1091 "learning_rate": 7.037468481535567e-05,1092 "loss": 0.5212,1093 "mean_token_accuracy": 0.8371426187455654,1094 "num_tokens": 527940592.0,1095 "step": 5751096 },1097 {1098 "epoch": 0.8246689771616458,1099 "grad_norm": 0.3154368910279167,1100 "learning_rate": 6.988699458685537e-05,1101 "loss": 0.5275,1102 "mean_token_accuracy": 0.8351783238351345,1103 "num_tokens": 532516910.0,1104 "step": 5801105 },1106 {1107 "epoch": 0.8317781924820048,1108 "grad_norm": 0.26226153440650185,1109 "learning_rate": 6.9397332949563e-05,1110 "loss": 0.5335,1111 "mean_token_accuracy": 0.8329351760447026,1112 "num_tokens": 537121758.0,1113 "step": 5851114 },1115 {1116 "epoch": 0.8388874078023638,1117 "grad_norm": 0.31223173870328286,1118 "learning_rate": 6.890576474687263e-05,1119 "loss": 0.5458,1120 "mean_token_accuracy": 0.829648780822754,1121 "num_tokens": 541734519.0,1122 "step": 5901123 },1124 {1125 "epoch": 0.8459966231227228,1126 "grad_norm": 0.2565970150956528,1127 "learning_rate": 6.841235507465515e-05,1128 "loss": 0.5415,1129 "mean_token_accuracy": 0.8324811846017838,1130 "num_tokens": 546326546.0,1131 "step": 5951132 },1133 {1134 "epoch": 0.8531058384430819,1135 "grad_norm": 0.29462309278409743,1136 "learning_rate": 6.791716927263778e-05,1137 "loss": 0.5354,1138 "mean_token_accuracy": 0.8325764186680317,1139 "num_tokens": 550923667.0,1140 "step": 6001141 },1142 {1143 "epoch": 0.8531058384430819,1144 "eval_loss": 0.5030205249786377,1145 "eval_mean_token_accuracy": 0.8328834866222582,1146 "eval_num_tokens": 550923667.0,1147 "eval_runtime": 145.5099,1148 "eval_samples_per_second": 25.002,1149 "eval_steps_per_second": 0.783,1150 "step": 6001151 },1152 {1153 "epoch": 0.8602150537634409,1154 "grad_norm": 0.2995740161508053,1155 "learning_rate": 6.742027291575156e-05,1156 "loss": 0.5351,1157 "mean_token_accuracy": 0.8337548352777958,1158 "num_tokens": 555521300.0,1159 "step": 6051160 },1161 {1162 "epoch": 0.8673242690837999,1163 "grad_norm": 0.256895454866442,1164 "learning_rate": 6.692173180544768e-05,1165 "loss": 0.527,1166 "mean_token_accuracy": 0.8346491247415543,1167 "num_tokens": 560114622.0,1168 "step": 6101169 },1170 {1171 "epoch": 0.8744334844041589,1172 "grad_norm": 0.26124663621839667,1173 "learning_rate": 6.642161196098351e-05,1174 "loss": 0.5299,1175 "mean_token_accuracy": 0.835064522176981,1176 "num_tokens": 564707120.0,1177 "step": 6151178 },1179 {1180 "epoch": 0.8815426997245179,1181 "grad_norm": 0.30629789668279445,1182 "learning_rate": 6.591997961068024e-05,1183 "loss": 0.5391,1184 "mean_token_accuracy": 0.8325687229633332,1185 "num_tokens": 569285949.0,1186 "step": 6201187 },1188 {1189 "epoch": 0.888651915044877,1190 "grad_norm": 0.2517010032545197,1191 "learning_rate": 6.541690118315245e-05,1192 "loss": 0.528,1193 "mean_token_accuracy": 0.834906804561615,1194 "num_tokens": 573871769.0,1195 "step": 6251196 },1197 {1198 "epoch": 0.895761130365236,1199 "grad_norm": 0.3714356282666368,1200 "learning_rate": 6.491244329851133e-05,