tttx/model_step2_3augs_buffer_short
05
1{2 "best_metric": null,3 "best_model_checkpoint": null,4 "epoch": 2.0,5 "eval_steps": 500,6 "global_step": 76,7 "is_hyper_param_search": false,8 "is_local_process_zero": true,9 "is_world_process_zero": true,10 "log_history": [11 {12 "epoch": 0.02631578947368421,13 "grad_norm": 0.011518988620826731,14 "learning_rate": 1e-05,15 "loss": 0.1758,16 "step": 117 },18 {19 "epoch": 0.05263157894736842,20 "grad_norm": 0.011769415675267096,21 "learning_rate": 2e-05,22 "loss": 0.1755,23 "step": 224 },25 {26 "epoch": 0.07894736842105263,27 "grad_norm": 0.012019294923116772,28 "learning_rate": 3.0000000000000004e-05,29 "loss": 0.1768,30 "step": 331 },32 {33 "epoch": 0.10526315789473684,34 "grad_norm": 0.011438929874712248,35 "learning_rate": 4e-05,36 "loss": 0.174,37 "step": 438 },39 {40 "epoch": 0.13157894736842105,41 "grad_norm": 0.012678409461035527,42 "learning_rate": 5e-05,43 "loss": 0.1776,44 "step": 545 },46 {47 "epoch": 0.15789473684210525,48 "grad_norm": 0.013302694501158712,49 "learning_rate": 6.000000000000001e-05,50 "loss": 0.1722,51 "step": 652 },53 {54 "epoch": 0.18421052631578946,55 "grad_norm": 0.015877560046165782,56 "learning_rate": 7.000000000000001e-05,57 "loss": 0.1717,58 "step": 759 },60 {61 "epoch": 0.21052631578947367,62 "grad_norm": 0.017316329988102695,63 "learning_rate": 8e-05,64 "loss": 0.1655,65 "step": 866 },67 {68 "epoch": 0.23684210526315788,69 "grad_norm": 0.01761728040990777,70 "learning_rate": 7.995731899209491e-05,71 "loss": 0.1615,72 "step": 973 },74 {75 "epoch": 0.2631578947368421,76 "grad_norm": 0.017422282759872943,77 "learning_rate": 7.982936705180139e-05,78 "loss": 0.1554,79 "step": 1080 },81 {82 "epoch": 0.2894736842105263,83 "grad_norm": 0.02050212432466269,84 "learning_rate": 7.961641723500821e-05,85 "loss": 0.1502,86 "step": 1187 },88 {89 "epoch": 0.3157894736842105,90 "grad_norm": 0.020386102383746066,91 "learning_rate": 7.931892398735608e-05,92 "loss": 0.1455,93 "step": 1294 },95 {96 "epoch": 0.34210526315789475,97 "grad_norm": 0.017501699031872063,98 "learning_rate": 7.893752217442773e-05,99 "loss": 0.1391,100 "step": 13101 },102 {103 "epoch": 0.3684210526315789,104 "grad_norm": 0.01644790453050047,105 "learning_rate": 7.847302572691277e-05,106 "loss": 0.1379,107 "step": 14108 },109 {110 "epoch": 0.39473684210526316,111 "grad_norm": 0.016205477891979105,112 "learning_rate": 7.792642590363864e-05,113 "loss": 0.1318,114 "step": 15115 },116 {117 "epoch": 0.42105263157894735,118 "grad_norm": 0.017772585025772586,119 "learning_rate": 7.729888917617424e-05,120 "loss": 0.1292,121 "step": 16122 },123 {124 "epoch": 0.4473684210526316,125 "grad_norm": 0.01437597741326583,126 "learning_rate": 7.659175473952085e-05,127 "loss": 0.1232,128 "step": 17129 },130 {131 "epoch": 0.47368421052631576,132 "grad_norm": 0.012880938282018417,133 "learning_rate": 7.58065316542025e-05,134 "loss": 0.119,135 "step": 18136 },137 {138 "epoch": 0.5,139 "grad_norm": 0.011093246487293535,140 "learning_rate": 7.494489562585479e-05,141 "loss": 0.1172,142 "step": 19143 },144 {145 "epoch": 0.5263157894736842,146 "grad_norm": 0.010467102848930952,147 "learning_rate": 7.400868542918457e-05,148 "loss": 0.1114,149 "step": 20150 },151 {152 "epoch": 0.5526315789473685,153 "grad_norm": 0.010047970328554336,154 "learning_rate": 7.299989898393209e-05,155 "loss": 0.1072,156 "step": 21157 },158 {159 "epoch": 0.5789473684210527,160 "grad_norm": 0.009892044194784507,161 "learning_rate": 7.192068909120959e-05,162 "loss": 0.1026,163 "step": 22164 },165 {166 "epoch": 0.6052631578947368,167 "grad_norm": 0.00977895435289584,168 "learning_rate": 7.077335883931516e-05,169 "loss": 0.1003,170 "step": 23171 },172 {173 "epoch": 0.631578947368421,174 "grad_norm": 0.009536071160931239,175 "learning_rate": 6.956035668882637e-05,176 "loss": 0.0963,177 "step": 24178 },179 {180 "epoch": 0.6578947368421053,181 "grad_norm": 0.009733333762799204,182 "learning_rate": 6.828427124746191e-05,183 "loss": 0.0913,184 "step": 25185 },186 {187 "epoch": 0.6842105263157895,188 "grad_norm": 0.009570441521026768,189 "learning_rate": 6.69478257458623e-05,190 "loss": 0.0891,191 "step": 26192 },193 {194 "epoch": 0.7105263157894737,195 "grad_norm": 0.01000020090090365,196 "learning_rate": 6.555387222607845e-05,197 "loss": 0.0829,198 "step": 27199 },200 {201 "epoch": 0.7368421052631579,202 "grad_norm": 0.00996936468337858,203 "learning_rate": 6.410538545517026e-05,204 "loss": 0.0807,205 "step": 28206 },207 {208 "epoch": 0.7631578947368421,209 "grad_norm": 0.00953954885336217,210 "learning_rate": 6.260545657690369e-05,211 "loss": 0.0805,212 "step": 29213 },214 {215 "epoch": 0.7894736842105263,216 "grad_norm": 0.009088232136551395,217 "learning_rate": 6.105728651509424e-05,218 "loss": 0.0768,219 "step": 30220 },221 {222 "epoch": 0.8157894736842105,223 "grad_norm": 0.008652738141347285,224 "learning_rate": 5.946417914267425e-05,225 "loss": 0.074,226 "step": 31227 },228 {229 "epoch": 0.8421052631578947,230 "grad_norm": 0.007369187149441024,231 "learning_rate": 5.782953423106154e-05,232 "loss": 0.0726,233 "step": 32234 },235 {236 "epoch": 0.868421052631579,237 "grad_norm": 0.007130681951299134,238 "learning_rate": 5.61568401948758e-05,239 "loss": 0.0719,240 "step": 33241 },242 {243 "epoch": 0.8947368421052632,244 "grad_norm": 0.007528001216497669,245 "learning_rate": 5.444966664748613e-05,246 "loss": 0.0706,247 "step": 34248 },249 {250 "epoch": 0.9210526315789473,251 "grad_norm": 0.008072156901279035,252 "learning_rate": 5.271165678327607e-05,253 "loss": 0.0707,254 "step": 35255 },256 {257 "epoch": 0.9473684210526315,258 "grad_norm": 0.007829567280354091,259 "learning_rate": 5.0946519602883326e-05,260 "loss": 0.0669,261 "step": 36262 },263 {264 "epoch": 0.9736842105263158,265 "grad_norm": 0.00799333236262416,266 "learning_rate": 4.9158021998005366e-05,267 "loss": 0.0661,268 "step": 37269 },270 {271 "epoch": 1.0,272 "grad_norm": 0.00816214803013346,273 "learning_rate": 4.734998071266282e-05,274 "loss": 0.063,275 "step": 38276 },277 {278 "epoch": 1.0263157894736843,279 "grad_norm": 0.007720443171687867,280 "learning_rate": 4.552625419807529e-05,281 "loss": 0.0608,282 "step": 39283 },284 {285 "epoch": 1.0526315789473684,286 "grad_norm": 0.008122893547240941,287 "learning_rate": 4.369073437853208e-05,288 "loss": 0.0608,289 "step": 40290 },291 {292 "epoch": 1.0789473684210527,293 "grad_norm": 0.008220632700161948,294 "learning_rate": 4.184733834582959e-05,295 "loss": 0.0586,296 "step": 41297 },298 {299 "epoch": 1.1052631578947367,300 "grad_norm": 0.008642996154857054,301 "learning_rate": 4e-05,302 "loss": 0.0618,303 "step": 42304 },305 {306 "epoch": 1.131578947368421,307 "grad_norm": 0.008332657542430085,308 "learning_rate": 3.815266165417043e-05,309 "loss": 0.0574,310 "step": 43311 },312 {313 "epoch": 1.1578947368421053,314 "grad_norm": 0.008567867343178283,315 "learning_rate": 3.630926562146792e-05,316 "loss": 0.0589,317 "step": 44318 },319 {320 "epoch": 1.1842105263157894,321 "grad_norm": 0.008265949758336088,322 "learning_rate": 3.447374580192473e-05,323 "loss": 0.0543,324 "step": 45325 },326 {327 "epoch": 1.2105263157894737,328 "grad_norm": 0.008431332834549176,329 "learning_rate": 3.2650019287337184e-05,330 "loss": 0.0536,331 "step": 46332 },333 {334 "epoch": 1.236842105263158,335 "grad_norm": 0.007829990670514665,336 "learning_rate": 3.084197800199465e-05,337 "loss": 0.0531,338 "step": 47339 },340 {341 "epoch": 1.263157894736842,342 "grad_norm": 0.00805463741954765,343 "learning_rate": 2.905348039711669e-05,344 "loss": 0.0532,345 "step": 48346 },347 {348 "epoch": 1.2894736842105263,349 "grad_norm": 0.007944851951870653,350 "learning_rate": 2.728834321672394e-05,351 "loss": 0.0529,352 "step": 49353 },354 {355 "epoch": 1.3157894736842106,356 "grad_norm": 0.00787719509680645,357 "learning_rate": 2.5550333352513885e-05,358 "loss": 0.051,359 "step": 50360 },361 {362 "epoch": 1.3421052631578947,363 "grad_norm": 0.007689730911636538,364 "learning_rate": 2.3843159805124207e-05,365 "loss": 0.0499,366 "step": 51367 },368 {369 "epoch": 1.368421052631579,370 "grad_norm": 0.008145577867986397,371 "learning_rate": 2.2170465768938473e-05,372 "loss": 0.0497,373 "step": 52374 },375 {376 "epoch": 1.3947368421052633,377 "grad_norm": 0.007765854028854914,378 "learning_rate": 2.0535820857325755e-05,379 "loss": 0.0488,380 "step": 53381 },382 {383 "epoch": 1.4210526315789473,384 "grad_norm": 0.010541614489525645,385 "learning_rate": 1.8942713484905762e-05,386 "loss": 0.05,387 "step": 54388 },389 {390 "epoch": 1.4473684210526316,391 "grad_norm": 0.009377352769384973,392 "learning_rate": 1.7394543423096327e-05,393 "loss": 0.0504,394 "step": 55395 },396 {397 "epoch": 1.4736842105263157,398 "grad_norm": 0.007879030184576382,399 "learning_rate": 1.589461454482975e-05,400 "loss": 0.0485,401 "step": 56402 },403 {404 "epoch": 1.5,405 "grad_norm": 0.008528311766387752,406 "learning_rate": 1.4446127773921559e-05,407 "loss": 0.0506,408 "step": 57409 },410 {411 "epoch": 1.526315789473684,412 "grad_norm": 0.0076161967841271135,413 "learning_rate": 1.3052174254137713e-05,414 "loss": 0.0486,415 "step": 58416 },417 {418 "epoch": 1.5526315789473686,419 "grad_norm": 0.0069022846350392285,420 "learning_rate": 1.1715728752538103e-05,421 "loss": 0.0469,422 "step": 59423 },424 {425 "epoch": 1.5789473684210527,426 "grad_norm": 0.00963107640357555,427 "learning_rate": 1.0439643311173642e-05,428 "loss": 0.0488,429 "step": 60430 },431 {432 "epoch": 1.6052631578947367,433 "grad_norm": 0.006604996990410402,434 "learning_rate": 9.226641160684843e-06,435 "loss": 0.0484,436 "step": 61437 },438 {439 "epoch": 1.631578947368421,440 "grad_norm": 0.0059879956487925235,441 "learning_rate": 8.07931090879042e-06,442 "loss": 0.0461,443 "step": 62444 },445 {446 "epoch": 1.6578947368421053,447 "grad_norm": 0.0061371313058644255,448 "learning_rate": 7.000101016067913e-06,449 "loss": 0.0454,450 "step": 63451 },452 {453 "epoch": 1.6842105263157894,454 "grad_norm": 0.007073804479577813,455 "learning_rate": 5.991314570815441e-06,456 "loss": 0.049,457 "step": 64458 },459 {460 "epoch": 1.7105263157894737,461 "grad_norm": 0.006925567210585072,462 "learning_rate": 5.055104374145221e-06,463 "loss": 0.0471,464 "step": 65465 },466 {467 "epoch": 1.736842105263158,468 "grad_norm": 0.00679527298679942,469 "learning_rate": 4.193468345797511e-06,470 "loss": 0.0464,471 "step": 66472 },473 {474 "epoch": 1.763157894736842,475 "grad_norm": 0.005915070559407378,476 "learning_rate": 3.4082452604791594e-06,477 "loss": 0.0449,478 "step": 67479 },480 {481 "epoch": 1.7894736842105263,482 "grad_norm": 0.0065205977736393144,483 "learning_rate": 2.7011108238257723e-06,484 "loss": 0.0471,485 "step": 68486 },487 {488 "epoch": 1.8157894736842106,489 "grad_norm": 0.006260173837861193,490 "learning_rate": 2.073574096361366e-06,491 "loss": 0.0486,492 "step": 69493 },494 {495 "epoch": 1.8421052631578947,496 "grad_norm": 0.006245995730039592,497 "learning_rate": 1.5269742730872384e-06,498 "loss": 0.0473,499 "step": 70500 },501 {502 "epoch": 1.868421052631579,503 "grad_norm": 0.00654850126698062,504 "learning_rate": 1.0624778255722857e-06,505 "loss": 0.0477,506 "step": 71507 },508 {509 "epoch": 1.8947368421052633,510 "grad_norm": 0.008481528591018958,511 "learning_rate": 6.810760126439287e-07,512 "loss": 0.0488,513 "step": 72514 },515 {516 "epoch": 1.9210526315789473,517 "grad_norm": 0.006002865737944914,518 "learning_rate": 3.835827649917967e-07,519 "loss": 0.0456,520 "step": 73521 },522 {523 "epoch": 1.9473684210526314,524 "grad_norm": 0.0061084652605397196,525 "learning_rate": 1.706329481986213e-07,526 "loss": 0.0484,527 "step": 74528 },529 {530 "epoch": 1.973684210526316,531 "grad_norm": 0.006251582720819486,532 "learning_rate": 4.268100790510321e-08,533 "loss": 0.0461,534 "step": 75535 },536 {537 "epoch": 2.0,538 "grad_norm": 0.007872514188147876,539 "learning_rate": 0.0,540 "loss": 0.0473,541 "step": 76542 },543 {544 "epoch": 2.0,545 "step": 76,546 "total_flos": 114338546712576.0,547 "train_loss": 0.08416653395091232,548 "train_runtime": 2655.6769,549 "train_samples_per_second": 0.451,550 "train_steps_per_second": 0.029551 }552 ],553 "logging_steps": 1,554 "max_steps": 76,555 "num_input_tokens_seen": 0,556 "num_train_epochs": 2,557 "save_steps": 500,558 "stateful_callbacks": {559 "TrainerControl": {560 "args": {561 "should_epoch_stop": false,562 "should_evaluate": false,563 "should_log": false,564 "should_save": false,565 "should_training_stop": false566 },567 "attributes": {}568 }569 },570 "total_flos": 114338546712576.0,571 "train_batch_size": 2,572 "trial_name": null,573 "trial_params": null574}575 