LocalLite/python-code-refactoring-ministral-8B
0
1{2 "best_global_step": 400,3 "best_metric": 0.2452806532382965,4 "best_model_checkpoint": "./python-code-refactoring-ministral-8B/checkpoint-400",5 "epoch": 1.88,6 "eval_steps": 50,7 "global_step": 400,8 "is_hyper_param_search": false,9 "is_local_process_zero": true,10 "is_world_process_zero": true,11 "log_history": [12 {13 "entropy": 1.2112591058015822,14 "epoch": 0.047058823529411764,15 "grad_norm": 0.6328933835029602,16 "learning_rate": 7.500000000000001e-06,17 "loss": 1.7004,18 "mean_token_accuracy": 0.6649438560009002,19 "num_tokens": 16836.0,20 "step": 1021 },22 {23 "entropy": 1.2078000918030738,24 "epoch": 0.09411764705882353,25 "grad_norm": 0.5308480858802795,26 "learning_rate": 9.991971098713136e-06,27 "loss": 1.5909,28 "mean_token_accuracy": 0.6793556809425354,29 "num_tokens": 33851.0,30 "step": 2031 },32 {33 "entropy": 1.193270905315876,34 "epoch": 0.1411764705882353,35 "grad_norm": 0.5148553252220154,36 "learning_rate": 9.952707929045018e-06,37 "loss": 1.3732,38 "mean_token_accuracy": 0.7048414140939713,39 "num_tokens": 51118.0,40 "step": 3041 },42 {43 "entropy": 1.1802570581436158,44 "epoch": 0.18823529411764706,45 "grad_norm": 0.5968846082687378,46 "learning_rate": 9.88099273969759e-06,47 "loss": 1.1662,48 "mean_token_accuracy": 0.728707967698574,49 "num_tokens": 67851.0,50 "step": 4051 },52 {53 "entropy": 0.8819854319095611,54 "epoch": 0.23529411764705882,55 "grad_norm": 0.5468736290931702,56 "learning_rate": 9.777295435775165e-06,57 "loss": 0.8373,58 "mean_token_accuracy": 0.815536054968834,59 "num_tokens": 84988.0,60 "step": 5061 },62 {63 "epoch": 0.23529411764705882,64 "eval_entropy": 0.7242053945859274,65 "eval_loss": 0.6968294978141785,66 "eval_mean_token_accuracy": 0.8591729084650676,67 "eval_num_tokens": 84988.0,68 "eval_runtime": 18.242,69 "eval_samples_per_second": 8.223,70 "eval_steps_per_second": 8.223,71 "step": 5072 },73 {74 "entropy": 0.593521349132061,75 "epoch": 0.2823529411764706,76 "grad_norm": 0.3674946427345276,77 "learning_rate": 9.64229548131959e-06,78 "loss": 0.5825,79 "mean_token_accuracy": 0.8873148798942566,80 "num_tokens": 102092.0,81 "step": 6082 },83 {84 "entropy": 0.45081270560622216,85 "epoch": 0.32941176470588235,86 "grad_norm": 0.40957480669021606,87 "learning_rate": 9.476877447204308e-06,88 "loss": 0.4838,89 "mean_token_accuracy": 0.9021970316767692,90 "num_tokens": 118935.0,91 "step": 7092 },93 {94 "entropy": 0.36085937917232513,95 "epoch": 0.3764705882352941,96 "grad_norm": 0.39999279379844666,97 "learning_rate": 9.282125215090694e-06,98 "loss": 0.3868,99 "mean_token_accuracy": 0.919852776825428,100 "num_tokens": 135700.0,101 "step": 80102 },103 {104 "entropy": 0.3600465551018715,105 "epoch": 0.4235294117647059,106 "grad_norm": 0.24208979308605194,107 "learning_rate": 9.059314875424553e-06,108 "loss": 0.3347,109 "mean_token_accuracy": 0.9249877437949181,110 "num_tokens": 152723.0,111 "step": 90112 },113 {114 "entropy": 0.3371710181236267,115 "epoch": 0.47058823529411764,116 "grad_norm": 0.24354608356952667,117 "learning_rate": 8.809906366007832e-06,118 "loss": 0.3068,119 "mean_token_accuracy": 0.9299718767404557,120 "num_tokens": 169857.0,121 "step": 100122 },123 {124 "epoch": 0.47058823529411764,125 "eval_entropy": 0.3274345314502716,126 "eval_loss": 0.3101823031902313,127 "eval_mean_token_accuracy": 0.9297333884239197,128 "eval_num_tokens": 169857.0,129 "eval_runtime": 18.3304,130 "eval_samples_per_second": 8.183,131 "eval_steps_per_second": 8.183,132 "step": 100133 },134 {135 "entropy": 0.3301278352737427,136 "epoch": 0.5176470588235295,137 "grad_norm": 0.24222862720489502,138 "learning_rate": 8.535533905932739e-06,139 "loss": 0.297,140 "mean_token_accuracy": 0.9317866399884224,141 "num_tokens": 186799.0,142 "step": 110143 },144 {145 "entropy": 0.3344870314002037,146 "epoch": 0.5647058823529412,147 "grad_norm": 0.24282677471637726,148 "learning_rate": 8.237995287558801e-06,149 "loss": 0.2942,150 "mean_token_accuracy": 0.9307269603013992,151 "num_tokens": 203775.0,152 "step": 120153 },154 {155 "entropy": 0.32817855961620807,156 "epoch": 0.611764705882353,157 "grad_norm": 0.20758755505084991,158 "learning_rate": 7.919240096695877e-06,159 "loss": 0.2832,160 "mean_token_accuracy": 0.9334745943546295,161 "num_tokens": 220919.0,162 "step": 130163 },164 {165 "entropy": 0.2978081665933132,166 "epoch": 0.6588235294117647,167 "grad_norm": 0.25893646478652954,168 "learning_rate": 7.58135693817893e-06,169 "loss": 0.2731,170 "mean_token_accuracy": 0.9333468064665794,171 "num_tokens": 238215.0,172 "step": 140173 },174 {175 "entropy": 0.28119575753808024,176 "epoch": 0.7058823529411765,177 "grad_norm": 0.26739779114723206,178 "learning_rate": 7.226559750537462e-06,179 "loss": 0.2673,180 "mean_token_accuracy": 0.9358425229787827,181 "num_tokens": 254625.0,182 "step": 150183 },184 {185 "epoch": 0.7058823529411765,186 "eval_entropy": 0.2825607425967852,187 "eval_loss": 0.27234065532684326,188 "eval_mean_token_accuracy": 0.9340071042378744,189 "eval_num_tokens": 254625.0,190 "eval_runtime": 18.2949,191 "eval_samples_per_second": 8.199,192 "eval_steps_per_second": 8.199,193 "step": 150194 },195 {196 "entropy": 0.28226450830698013,197 "epoch": 0.7529411764705882,198 "grad_norm": 0.27276191115379333,199 "learning_rate": 6.857173299431084e-06,200 "loss": 0.2697,201 "mean_token_accuracy": 0.9357088133692741,202 "num_tokens": 271437.0,203 "step": 160204 },205 {206 "entropy": 0.2687517054378986,207 "epoch": 0.8,208 "grad_norm": 0.2764045000076294,209 "learning_rate": 6.475617944903693e-06,210 "loss": 0.2438,211 "mean_token_accuracy": 0.9380499675869942,212 "num_tokens": 288419.0,213 "step": 170214 },215 {216 "entropy": 0.2816390782594681,217 "epoch": 0.8470588235294118,218 "grad_norm": 0.23346015810966492,219 "learning_rate": 6.084393782267039e-06,220 "loss": 0.2607,221 "mean_token_accuracy": 0.9347616925835609,222 "num_tokens": 305535.0,223 "step": 180224 },225 {226 "entropy": 0.2635271079838276,227 "epoch": 0.8941176470588236,228 "grad_norm": 0.2241382598876953,229 "learning_rate": 5.686064260528577e-06,230 "loss": 0.2437,231 "mean_token_accuracy": 0.9385963633656502,232 "num_tokens": 322499.0,233 "step": 190234 },235 {236 "entropy": 0.2634481403976679,237 "epoch": 0.9411764705882353,238 "grad_norm": 0.25238925218582153,239 "learning_rate": 5.283239385701882e-06,240 "loss": 0.2478,241 "mean_token_accuracy": 0.9358779728412628,242 "num_tokens": 340031.0,243 "step": 200244 },245 {246 "epoch": 0.9411764705882353,247 "eval_entropy": 0.2671808664004008,248 "eval_loss": 0.25436246395111084,249 "eval_mean_token_accuracy": 0.9354545970757803,250 "eval_num_tokens": 340031.0,251 "eval_runtime": 18.3601,252 "eval_samples_per_second": 8.17,253 "eval_steps_per_second": 8.17,254 "step": 200255 },256 {257 "entropy": 0.27091246508061884,258 "epoch": 0.9882352941176471,259 "grad_norm": 0.2224678099155426,260 "learning_rate": 4.8785586190579355e-06,261 "loss": 0.2495,262 "mean_token_accuracy": 0.9346240475773812,263 "num_tokens": 356961.0,264 "step": 210265 },266 {267 "entropy": 0.27871040882248627,268 "epoch": 1.0329411764705883,269 "grad_norm": 0.24531640112400055,270 "learning_rate": 4.474673582374381e-06,271 "loss": 0.2539,272 "mean_token_accuracy": 0.9362887332313939,273 "num_tokens": 372730.0,274 "step": 220275 },276 {277 "entropy": 0.256845248490572,278 "epoch": 1.08,279 "grad_norm": 0.21706384420394897,280 "learning_rate": 4.074230683504541e-06,281 "loss": 0.2392,282 "mean_token_accuracy": 0.9392171293497086,283 "num_tokens": 389960.0,284 "step": 230285 },286 {287 "entropy": 0.2579733207821846,288 "epoch": 1.1270588235294117,289 "grad_norm": 0.19243040680885315,290 "learning_rate": 3.679853776110035e-06,291 "loss": 0.2459,292 "mean_token_accuracy": 0.9344916313886642,293 "num_tokens": 406947.0,294 "step": 240295 },296 {297 "entropy": 0.26577013581991193,298 "epoch": 1.1741176470588235,299 "grad_norm": 0.2271624207496643,300 "learning_rate": 3.2941269671770193e-06,301 "loss": 0.2552,302 "mean_token_accuracy": 0.9344739750027656,303 "num_tokens": 423908.0,304 "step": 250305 },306 {307 "epoch": 1.1741176470588235,308 "eval_entropy": 0.2570107207695643,309 "eval_loss": 0.2488705962896347,310 "eval_mean_token_accuracy": 0.9355408449967703,311 "eval_num_tokens": 423908.0,312 "eval_runtime": 18.344,313 "eval_samples_per_second": 8.177,314 "eval_steps_per_second": 8.177,315 "step": 250316 },317 {318 "entropy": 0.2559947054833174,319 "epoch": 1.2211764705882353,320 "grad_norm": 0.24396584928035736,321 "learning_rate": 2.9195776849677037e-06,322 "loss": 0.2448,323 "mean_token_accuracy": 0.9389567241072655,324 "num_tokens": 440658.0,325 "step": 260326 },327 {328 "entropy": 0.2555523537099361,329 "epoch": 1.2682352941176471,330 "grad_norm": 0.21874727308750153,331 "learning_rate": 2.5586601183523485e-06,332 "loss": 0.2428,333 "mean_token_accuracy": 0.9377847194671631,334 "num_tokens": 457473.0,335 "step": 270336 },337 {338 "entropy": 0.2418661404401064,339 "epoch": 1.315294117647059,340 "grad_norm": 0.2622573971748352,341 "learning_rate": 2.213739136033533e-06,342 "loss": 0.2251,343 "mean_token_accuracy": 0.9425412893295289,344 "num_tokens": 474051.0,345 "step": 280346 },347 {348 "entropy": 0.2569988567382097,349 "epoch": 1.3623529411764705,350 "grad_norm": 0.21504490077495575,351 "learning_rate": 1.8870747910300064e-06,352 "loss": 0.2421,353 "mean_token_accuracy": 0.9356839284300804,354 "num_tokens": 491292.0,355 "step": 290356 },357 {358 "entropy": 0.25272532626986505,359 "epoch": 1.4094117647058824,360 "grad_norm": 0.3225324749946594,361 "learning_rate": 1.5808075119526323e-06,362 "loss": 0.2415,363 "mean_token_accuracy": 0.9385002166032791,364 "num_tokens": 508173.0,365 "step": 300366 },367 {368 "epoch": 1.4094117647058824,369 "eval_entropy": 0.2538743743300438,370 "eval_loss": 0.24622248113155365,371 "eval_mean_token_accuracy": 0.936361878712972,372 "eval_num_tokens": 508173.0,373 "eval_runtime": 18.3721,374 "eval_samples_per_second": 8.165,375 "eval_steps_per_second": 8.165,376 "step": 300377 },378 {379 "entropy": 0.2525585234165192,380 "epoch": 1.4564705882352942,381 "grad_norm": 0.2269955575466156,382 "learning_rate": 1.2969440781048015e-06,383 "loss": 0.2353,384 "mean_token_accuracy": 0.9391644984483719,385 "num_tokens": 525115.0,386 "step": 310387 },388 {389 "entropy": 0.24930565245449543,390 "epoch": 1.5035294117647058,391 "grad_norm": 0.2441737949848175,392 "learning_rate": 1.0373444703037644e-06,393 "loss": 0.2333,394 "mean_token_accuracy": 0.9403112232685089,395 "num_tokens": 541969.0,396 "step": 320397 },398 {399 "entropy": 0.2371720265597105,400 "epoch": 1.5505882352941176,401 "grad_norm": 0.26385828852653503,402 "learning_rate": 8.037096835812885e-07,403 "loss": 0.2195,404 "mean_token_accuracy": 0.9421997696161271,405 "num_tokens": 558705.0,406 "step": 330407 },408 {409 "entropy": 0.2542194213718176,410 "epoch": 1.5976470588235294,411 "grad_norm": 0.2517501711845398,412 "learning_rate": 5.97570581619446e-07,413 "loss": 0.2382,414 "mean_token_accuracy": 0.9359523132443428,415 "num_tokens": 575605.0,416 "step": 340417 },418 {419 "entropy": 0.2532818391919136,420 "epoch": 1.644705882352941,421 "grad_norm": 0.23593880236148834,422 "learning_rate": 4.2027786595149555e-07,423 "loss": 0.2327,424 "mean_token_accuracy": 0.9397944405674934,425 "num_tokens": 592558.0,426 "step": 350427 },428 {429 "epoch": 1.644705882352941,430 "eval_entropy": 0.25020226001739504,431 "eval_loss": 0.2454231232404709,432 "eval_mean_token_accuracy": 0.9366115434964498,433 "eval_num_tokens": 592558.0,434 "eval_runtime": 18.3239,435 "eval_samples_per_second": 8.186,436 "eval_steps_per_second": 8.186,437 "step": 350438 },439 {440 "entropy": 0.25380014926195144,441 "epoch": 1.691764705882353,442 "grad_norm": 0.2268085479736328,443 "learning_rate": 2.7299322565344956e-07,444 "loss": 0.235,445 "mean_token_accuracy": 0.9367626756429672,446 "num_tokens": 609468.0,447 "step": 360448 },449 {450 "entropy": 0.24986149594187737,451 "epoch": 1.7388235294117647,452 "grad_norm": 0.23303046822547913,453 "learning_rate": 1.5668172551691174e-07,454 "loss": 0.2422,455 "mean_token_accuracy": 0.9362318605184555,456 "num_tokens": 626324.0,457 "step": 370458 },459 {460 "entropy": 0.2423665601760149,461 "epoch": 1.7858823529411765,462 "grad_norm": 0.23102161288261414,463 "learning_rate": 7.210548257874927e-08,464 "loss": 0.2218,465 "mean_token_accuracy": 0.9416185393929482,466 "num_tokens": 643181.0,467 "step": 380468 },469 {470 "entropy": 0.24225412756204606,471 "epoch": 1.8329411764705883,472 "grad_norm": 0.20650127530097961,473 "learning_rate": 1.981867244139124e-08,474 "loss": 0.2307,475 "mean_token_accuracy": 0.9393002942204476,476 "num_tokens": 660451.0,477 "step": 390478 },479 {480 "entropy": 0.24758095815777778,481 "epoch": 1.88,482 "grad_norm": 0.21503891050815582,483 "learning_rate": 1.6389810421846286e-10,484 "loss": 0.2379,485 "mean_token_accuracy": 0.9382512927055359,486 "num_tokens": 677476.0,487 "step": 400488 },489 {490 "epoch": 1.88,491 "eval_entropy": 0.24997014413277308,492 "eval_loss": 0.2452806532382965,493 "eval_mean_token_accuracy": 0.9366318229834238,494 "eval_num_tokens": 677476.0,495 "eval_runtime": 18.3789,496 "eval_samples_per_second": 8.162,497 "eval_steps_per_second": 8.162,498 "step": 400499 }500 ],501 "logging_steps": 10,502 "max_steps": 400,503 "num_input_tokens_seen": 0,504 "num_train_epochs": 2,505 "save_steps": 100,506 "stateful_callbacks": {507 "TrainerControl": {508 "args": {509 "should_epoch_stop": false,510 "should_evaluate": false,511 "should_log": false,512 "should_save": true,513 "should_training_stop": true514 },515 "attributes": {}516 }517 },518 "total_flos": 3.112673744948429e+16,519 "train_batch_size": 1,520 "trial_name": null,521 "trial_params": null522}523