LocalLite/python-code-completion-ministral-8B
0
1{2 "best_global_step": 150,3 "best_metric": 0.5928216576576233,4 "best_model_checkpoint": "./python-code-completion-ministral-8B/checkpoint-100",5 "epoch": 3.9098039215686273,6 "eval_steps": 50,7 "global_step": 500,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "entropy": 1.5072221472859382,14 "epoch": 0.0784313725490196,15 "grad_norm": 0.355531245470047,16 "learning_rate": 6e-05,17 "loss": 1.7279,18 "mean_token_accuracy": 0.6269892059266567,19 "num_tokens": 22720.0,20 "step": 1021 },22 {23 "entropy": 1.156855022162199,24 "epoch": 0.1568627450980392,25 "grad_norm": 0.4324820339679718,26 "learning_rate": 9.998321768134102e-05,27 "loss": 1.027,28 "mean_token_accuracy": 0.7633639872074127,29 "num_tokens": 46930.0,30 "step": 2031 },32 {33 "entropy": 0.724648866429925,34 "epoch": 0.23529411764705882,35 "grad_norm": 0.28367722034454346,36 "learning_rate": 9.979454595377594e-05,37 "loss": 0.7082,38 "mean_token_accuracy": 0.8542310416698455,39 "num_tokens": 73605.0,40 "step": 3041 },42 {43 "entropy": 0.6228405788540841,44 "epoch": 0.3137254901960784,45 "grad_norm": 0.2075912207365036,46 "learning_rate": 9.93970185956522e-05,47 "loss": 0.653,48 "mean_token_accuracy": 0.8736839070916176,49 "num_tokens": 98799.0,50 "step": 4051 },52 {53 "entropy": 0.5743573283776641,54 "epoch": 0.39215686274509803,55 "grad_norm": 0.21785929799079895,56 "learning_rate": 9.879230297486034e-05,57 "loss": 0.6108,58 "mean_token_accuracy": 0.8821258693933487,59 "num_tokens": 121501.0,60 "step": 5061 },62 {63 "epoch": 0.39215686274509803,64 "eval_entropy": 0.6596304257710774,65 "eval_loss": 0.6177071332931519,66 "eval_mean_token_accuracy": 0.8652889298068153,67 "eval_num_tokens": 121501.0,68 "eval_runtime": 15.4707,69 "eval_samples_per_second": 5.817,70 "eval_steps_per_second": 5.817,71 "step": 5072 },73 {74 "entropy": 0.6043186411261559,75 "epoch": 0.47058823529411764,76 "grad_norm": 0.31095993518829346,77 "learning_rate": 9.798293547886748e-05,78 "loss": 0.645,79 "mean_token_accuracy": 0.8754804000258446,80 "num_tokens": 145419.0,81 "step": 6082 },83 {84 "entropy": 0.6452899970114231,85 "epoch": 0.5490196078431373,86 "grad_norm": 0.3375164866447449,87 "learning_rate": 9.697231087622691e-05,88 "loss": 0.6634,89 "mean_token_accuracy": 0.8640451654791832,90 "num_tokens": 169873.0,91 "step": 7092 },93 {94 "entropy": 0.6401195958256721,95 "epoch": 0.6274509803921569,96 "grad_norm": 0.21431401371955872,97 "learning_rate": 9.576466807773899e-05,98 "loss": 0.6615,99 "mean_token_accuracy": 0.8669946551322937,100 "num_tokens": 194305.0,101 "step": 80102 },103 {104 "entropy": 0.5998968390747905,105 "epoch": 0.7058823529411765,106 "grad_norm": 0.22246505320072174,107 "learning_rate": 9.436507235698612e-05,108 "loss": 0.6405,109 "mean_token_accuracy": 0.877622701227665,110 "num_tokens": 217799.0,111 "step": 90112 },113 {114 "entropy": 0.5195517055690289,115 "epoch": 0.7843137254901961,116 "grad_norm": 0.22644591331481934,117 "learning_rate": 9.277939410481507e-05,118 "loss": 0.574,119 "mean_token_accuracy": 0.8936989307403564,120 "num_tokens": 241897.0,121 "step": 100122 },123 {124 "epoch": 0.7843137254901961,125 "eval_entropy": 0.607519467257791,126 "eval_loss": 0.5936470627784729,127 "eval_mean_token_accuracy": 0.8691287775834401,128 "eval_num_tokens": 241897.0,129 "eval_runtime": 15.5316,130 "eval_samples_per_second": 5.795,131 "eval_steps_per_second": 5.795,132 "step": 100133 },134 {135 "entropy": 0.598377226293087,136 "epoch": 0.8627450980392157,137 "grad_norm": 0.19417309761047363,138 "learning_rate": 9.101428420687759e-05,139 "loss": 0.6663,140 "mean_token_accuracy": 0.8794096082448959,141 "num_tokens": 266761.0,142 "step": 110143 },144 {145 "entropy": 0.5745990440249443,146 "epoch": 0.9411764705882353,147 "grad_norm": 0.30245068669319153,148 "learning_rate": 8.907714614750473e-05,149 "loss": 0.605,150 "mean_token_accuracy": 0.8777685552835465,151 "num_tokens": 291812.0,152 "step": 120153 },154 {155 "entropy": 0.563015444890449,156 "epoch": 1.0156862745098039,157 "grad_norm": 0.24953527748584747,158 "learning_rate": 8.697610495692054e-05,159 "loss": 0.5665,160 "mean_token_accuracy": 0.8813700503424594,161 "num_tokens": 314476.0,162 "step": 130163 },164 {165 "entropy": 0.53117360714823,166 "epoch": 1.0941176470588236,167 "grad_norm": 0.24786317348480225,168 "learning_rate": 8.471997313204181e-05,169 "loss": 0.5683,170 "mean_token_accuracy": 0.8960023656487465,171 "num_tokens": 335438.0,172 "step": 140173 },174 {175 "entropy": 0.48921620547771455,176 "epoch": 1.1725490196078432,177 "grad_norm": 0.33822259306907654,178 "learning_rate": 8.231821367380335e-05,179 "loss": 0.5322,180 "mean_token_accuracy": 0.8975507125258446,181 "num_tokens": 361329.0,182 "step": 150183 },184 {185 "epoch": 1.1725490196078432,186 "eval_entropy": 0.5694952074852254,187 "eval_loss": 0.5928216576576233,188 "eval_mean_token_accuracy": 0.8691950864262051,189 "eval_num_tokens": 361329.0,190 "eval_runtime": 15.4486,191 "eval_samples_per_second": 5.826,192 "eval_steps_per_second": 5.826,193 "step": 150194 },195 {196 "entropy": 0.5561734782531857,197 "epoch": 1.2509803921568627,198 "grad_norm": 0.27892613410949707,199 "learning_rate": 7.978090039604341e-05,200 "loss": 0.5711,201 "mean_token_accuracy": 0.8858156308531762,202 "num_tokens": 387834.0,203 "step": 160204 },205 {206 "entropy": 0.48290685191750526,207 "epoch": 1.3294117647058823,208 "grad_norm": 0.2847723662853241,209 "learning_rate": 7.711867567242768e-05,210 "loss": 0.5477,211 "mean_token_accuracy": 0.8980041235685349,212 "num_tokens": 412449.0,213 "step": 170214 },215 {216 "entropy": 0.4553432559594512,217 "epoch": 1.4078431372549018,218 "grad_norm": 0.35990357398986816,219 "learning_rate": 7.434270579863548e-05,220 "loss": 0.4953,221 "mean_token_accuracy": 0.9054195120930671,222 "num_tokens": 434822.0,223 "step": 180224 },225 {226 "entropy": 0.4999146230518818,227 "epoch": 1.4862745098039216,228 "grad_norm": 0.23805052042007446,229 "learning_rate": 7.146463415703531e-05,230 "loss": 0.5029,231 "mean_token_accuracy": 0.8930570259690285,232 "num_tokens": 461303.0,233 "step": 190234 },235 {236 "entropy": 0.49880330096930264,237 "epoch": 1.5647058823529412,238 "grad_norm": 0.2987266480922699,239 "learning_rate": 6.84965323802926e-05,240 "loss": 0.5443,241 "mean_token_accuracy": 0.8915592297911644,242 "num_tokens": 487363.0,243 "step": 200244 },245 {246 "epoch": 1.5647058823529412,247 "eval_entropy": 0.5494673619667689,248 "eval_loss": 0.5963289737701416,249 "eval_mean_token_accuracy": 0.8694895923137664,250 "eval_num_tokens": 487363.0,251 "eval_runtime": 15.5114,252 "eval_samples_per_second": 5.802,253 "eval_steps_per_second": 5.802,254 "step": 200255 },256 {257 "entropy": 0.49026294946670534,258 "epoch": 1.6431372549019607,259 "grad_norm": 0.2591567039489746,260 "learning_rate": 6.545084971874738e-05,261 "loss": 0.5028,262 "mean_token_accuracy": 0.8963919445872307,263 "num_tokens": 512437.0,264 "step": 210265 },266 {267 "entropy": 0.5070990648120641,268 "epoch": 1.7215686274509805,269 "grad_norm": 0.3534194231033325,270 "learning_rate": 6.23403608239317e-05,271 "loss": 0.5434,272 "mean_token_accuracy": 0.8927687123417855,273 "num_tokens": 534856.0,274 "step": 220275 },276 {277 "entropy": 0.5167016156017781,278 "epoch": 1.8,279 "grad_norm": 0.26480528712272644,280 "learning_rate": 5.9178112167241805e-05,281 "loss": 0.5265,282 "mean_token_accuracy": 0.8922785848379136,283 "num_tokens": 560263.0,284 "step": 230285 },286 {287 "entropy": 0.4376767996698618,288 "epoch": 1.8784313725490196,289 "grad_norm": 0.2970808148384094,290 "learning_rate": 5.597736731850295e-05,291 "loss": 0.4587,292 "mean_token_accuracy": 0.9108576759696007,293 "num_tokens": 582717.0,294 "step": 240295 },296 {297 "entropy": 0.5208950355648995,298 "epoch": 1.9568627450980394,299 "grad_norm": 0.27791768312454224,300 "learning_rate": 5.2751551313948246e-05,301 "loss": 0.5336,302 "mean_token_accuracy": 0.8880411878228187,303 "num_tokens": 608651.0,304 "step": 250305 },306 {307 "epoch": 1.9568627450980394,308 "eval_entropy": 0.5718106091022491,309 "eval_loss": 0.5953241586685181,310 "eval_mean_token_accuracy": 0.8688194268279605,311 "eval_num_tokens": 608651.0,312 "eval_runtime": 15.4979,313 "eval_samples_per_second": 5.807,314 "eval_steps_per_second": 5.807,315 "step": 250316 },317 {318 "entropy": 0.45974215472999375,319 "epoch": 2.0313725490196077,320 "grad_norm": 0.32541951537132263,321 "learning_rate": 4.9514194346951144e-05,322 "loss": 0.4736,323 "mean_token_accuracy": 0.9053617276643452,324 "num_tokens": 629040.0,325 "step": 260326 },327 {328 "entropy": 0.4414994928985834,329 "epoch": 2.1098039215686275,330 "grad_norm": 0.372414231300354,331 "learning_rate": 4.627887501769231e-05,332 "loss": 0.4428,333 "mean_token_accuracy": 0.9085879683494568,334 "num_tokens": 654385.0,335 "step": 270336 },337 {338 "entropy": 0.3868284210562706,339 "epoch": 2.1882352941176473,340 "grad_norm": 0.27554434537887573,341 "learning_rate": 4.305916337979168e-05,342 "loss": 0.4041,343 "mean_token_accuracy": 0.9162766322493553,344 "num_tokens": 677869.0,345 "step": 280346 },347 {348 "entropy": 0.48472463674843314,349 "epoch": 2.2666666666666666,350 "grad_norm": 0.3065558969974518,351 "learning_rate": 3.9868564022787505e-05,352 "loss": 0.4702,353 "mean_token_accuracy": 0.9028230115771294,354 "num_tokens": 705954.0,355 "step": 290356 },357 {358 "entropy": 0.35362908374518154,359 "epoch": 2.3450980392156864,360 "grad_norm": 0.42765629291534424,361 "learning_rate": 3.672045942919474e-05,362 "loss": 0.3858,363 "mean_token_accuracy": 0.9247043713927269,364 "num_tokens": 729855.0,365 "step": 300366 },367 {368 "epoch": 2.3450980392156864,369 "eval_entropy": 0.4895121798747116,370 "eval_loss": 0.633377194404602,371 "eval_mean_token_accuracy": 0.8656294153796302,372 "eval_num_tokens": 729855.0,373 "eval_runtime": 15.4754,374 "eval_samples_per_second": 5.816,375 "eval_steps_per_second": 5.816,376 "step": 300377 },378 {379 "entropy": 0.3393667917698622,380 "epoch": 2.4235294117647057,381 "grad_norm": 0.3438664972782135,382 "learning_rate": 3.362805384372267e-05,383 "loss": 0.3374,384 "mean_token_accuracy": 0.9306267336010933,385 "num_tokens": 750942.0,386 "step": 310387 },388 {389 "entropy": 0.3776489123702049,390 "epoch": 2.5019607843137255,391 "grad_norm": 0.3191138207912445,392 "learning_rate": 3.0604317890083676e-05,393 "loss": 0.3675,394 "mean_token_accuracy": 0.9214938268065452,395 "num_tokens": 774956.0,396 "step": 320397 },398 {399 "entropy": 0.39363621212542055,400 "epoch": 2.5803921568627453,401 "grad_norm": 0.4947509169578552,402 "learning_rate": 2.7661934167689884e-05,403 "loss": 0.4188,404 "mean_token_accuracy": 0.916774469614029,405 "num_tokens": 800996.0,406 "step": 330407 },408 {409 "entropy": 0.3987773071974516,410 "epoch": 2.6588235294117646,411 "grad_norm": 0.4273108243942261,412 "learning_rate": 2.481324405642369e-05,413 "loss": 0.4063,414 "mean_token_accuracy": 0.9157314419746398,415 "num_tokens": 826362.0,416 "step": 340417 },418 {419 "entropy": 0.4490684296935797,420 "epoch": 2.7372549019607844,421 "grad_norm": 0.47922441363334656,422 "learning_rate": 2.207019595260154e-05,423 "loss": 0.4576,424 "mean_token_accuracy": 0.9035186737775802,425 "num_tokens": 853427.0,426 "step": 350427 },428 {429 "epoch": 2.7372549019607844,430 "eval_entropy": 0.5091634632812606,431 "eval_loss": 0.6345335245132446,432 "eval_mean_token_accuracy": 0.8654518723487854,433 "eval_num_tokens": 853427.0,434 "eval_runtime": 15.4411,435 "eval_samples_per_second": 5.829,436 "eval_steps_per_second": 5.829,437 "step": 350438 },439 {440 "entropy": 0.3651732888072729,441 "epoch": 2.8156862745098037,442 "grad_norm": 0.4229300618171692,443 "learning_rate": 1.9444295153247487e-05,444 "loss": 0.3517,445 "mean_token_accuracy": 0.9295069724321365,446 "num_tokens": 876031.0,447 "step": 360448 },449 {450 "entropy": 0.37069291807711124,451 "epoch": 2.8941176470588235,452 "grad_norm": 0.383728951215744,453 "learning_rate": 1.694655559887914e-05,454 "loss": 0.3875,455 "mean_token_accuracy": 0.9240497961640358,456 "num_tokens": 898756.0,457 "step": 370458 },459 {460 "entropy": 0.3721807461231947,461 "epoch": 2.9725490196078432,462 "grad_norm": 0.40181341767311096,463 "learning_rate": 1.4587453677213348e-05,464 "loss": 0.3983,465 "mean_token_accuracy": 0.9216682553291321,466 "num_tokens": 921618.0,467 "step": 380468 },469 {470 "entropy": 0.4022895936903201,471 "epoch": 3.0470588235294116,472 "grad_norm": 0.32404816150665283,473 "learning_rate": 1.2376884281555484e-05,474 "loss": 0.3566,475 "mean_token_accuracy": 0.9227047578284615,476 "num_tokens": 948594.0,477 "step": 390478 },479 {480 "entropy": 0.3646719891577959,481 "epoch": 3.1254901960784314,482 "grad_norm": 0.3212927281856537,483 "learning_rate": 1.0324119308178166e-05,484 "loss": 0.3489,485 "mean_token_accuracy": 0.9308759719133377,486 "num_tokens": 972074.0,487 "step": 400488 },489 {490 "epoch": 3.1254901960784314,491 "eval_entropy": 0.47637585144903927,492 "eval_loss": 0.661882758140564,493 "eval_mean_token_accuracy": 0.8632458435164557,494 "eval_num_tokens": 972074.0,495 "eval_runtime": 15.4948,496 "eval_samples_per_second": 5.808,497 "eval_steps_per_second": 5.808,498 "step": 400499 },500 {501 "entropy": 0.30484746433794496,502 "epoch": 3.203921568627451,503 "grad_norm": 0.3374861776828766,504 "learning_rate": 8.437768766765974e-06,505 "loss": 0.3161,506 "mean_token_accuracy": 0.9372083112597466,507 "num_tokens": 994846.0,508 "step": 410509 },510 {511 "entropy": 0.35498267617076634,512 "epoch": 3.2823529411764705,513 "grad_norm": 0.36716318130493164,514 "learning_rate": 6.725744667042777e-06,515 "loss": 0.3387,516 "mean_token_accuracy": 0.9290374144911766,517 "num_tokens": 1019726.0,518 "step": 420519 },520 {521 "entropy": 0.3321664972230792,522 "epoch": 3.3607843137254902,523 "grad_norm": 0.4375359117984772,524 "learning_rate": 5.195227833053274e-06,525 "loss": 0.323,526 "mean_token_accuracy": 0.9353578373789787,527 "num_tokens": 1044999.0,528 "step": 430529 },530 {531 "entropy": 0.33164924252778294,532 "epoch": 3.43921568627451,533 "grad_norm": 0.48313814401626587,534 "learning_rate": 3.852637784291424e-06,535 "loss": 0.346,536 "mean_token_accuracy": 0.9326006546616554,537 "num_tokens": 1068850.0,538 "step": 440539 },540 {541 "entropy": 0.29590620547533036,542 "epoch": 3.5176470588235293,543 "grad_norm": 0.5358214378356934,544 "learning_rate": 2.703605810004772e-06,545 "loss": 0.2704,546 "mean_token_accuracy": 0.9426496475934982,547 "num_tokens": 1092853.0,548 "step": 450549 },550 {551 "epoch": 3.5176470588235293,552 "eval_entropy": 0.46879171944326825,553 "eval_loss": 0.6711850762367249,554 "eval_mean_token_accuracy": 0.8622583528359731,555 "eval_num_tokens": 1092853.0,556 "eval_runtime": 15.462,557 "eval_samples_per_second": 5.821,558 "eval_steps_per_second": 5.821,559 "step": 450560 },561 {562 "entropy": 0.3179448990151286,563 "epoch": 3.596078431372549,564 "grad_norm": 0.4179091155529022,565 "learning_rate": 1.7529513496103323e-06,566 "loss": 0.3118,567 "mean_token_accuracy": 0.9393924847245216,568 "num_tokens": 1115370.0,569 "step": 460570 },571 {572 "entropy": 0.31651122458279135,573 "epoch": 3.674509803921569,574 "grad_norm": 0.4960501492023468,575 "learning_rate": 1.004661778290783e-06,576 "loss": 0.3162,577 "mean_token_accuracy": 0.9382667735219001,578 "num_tokens": 1137241.0,579 "step": 470580 },581 {582 "entropy": 0.3232851158827543,583 "epoch": 3.7529411764705882,584 "grad_norm": 0.39605653285980225,585 "learning_rate": 4.6187568255726124e-07,586 "loss": 0.3284,587 "mean_token_accuracy": 0.9355289787054062,588 "num_tokens": 1162001.0,589 "step": 480590 },591 {592 "entropy": 0.3265875181183219,593 "epoch": 3.831372549019608,594 "grad_norm": 0.6258050799369812,595 "learning_rate": 1.268696959267679e-07,596 "loss": 0.3199,597 "mean_token_accuracy": 0.9376031413674355,598 "num_tokens": 1186075.0,599 "step": 490600 },601 {602 "entropy": 0.36884093452245,603 "epoch": 3.9098039215686273,604 "grad_norm": 0.44404006004333496,605 "learning_rate": 1.0489499300603278e-09,606 "loss": 0.3421,607 "mean_token_accuracy": 0.9266003414988517,608 "num_tokens": 1213302.0,609 "step": 500610 },611 {612 "epoch": 3.9098039215686273,613 "eval_entropy": 0.47118989841805564,614 "eval_loss": 0.6726338863372803,615 "eval_mean_token_accuracy": 0.8621971819135877,616 "eval_num_tokens": 1213302.0,617 "eval_runtime": 15.5109,618 "eval_samples_per_second": 5.802,619 "eval_steps_per_second": 5.802,620 "step": 500621 }622 ],623 "logging_steps": 10,624 "max_steps": 500,625 "num_input_tokens_seen": 0,626 "num_train_epochs": 4,627 "save_steps": 100,628 "stateful_callbacks": {629 "TrainerControl": {630 "args": {631 "should_epoch_stop": false,632 "should_evaluate": false,633 "should_log": false,634 "should_save": true,635 "should_training_stop": true636 },637 "attributes": {}638 }639 },640 "total_flos": 5.574534419069338e+16,641 "train_batch_size": 1,642 "trial_name": null,643 "trial_params": null644}645