| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 740, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.033783783783783786, | |
| "grad_norm": 11.91859090228093, | |
| "learning_rate": 6.7567567567567575e-06, | |
| "loss": 0.9462, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06756756756756757, | |
| "grad_norm": 4.53680244078793, | |
| "learning_rate": 1.3513513513513515e-05, | |
| "loss": 0.4666, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10135135135135136, | |
| "grad_norm": 3.573903925081475, | |
| "learning_rate": 2.0270270270270273e-05, | |
| "loss": 0.3568, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.13513513513513514, | |
| "grad_norm": 3.832108029582817, | |
| "learning_rate": 2.702702702702703e-05, | |
| "loss": 0.264, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.16891891891891891, | |
| "grad_norm": 2.7780428812660225, | |
| "learning_rate": 3.3783783783783784e-05, | |
| "loss": 0.2717, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.20270270270270271, | |
| "grad_norm": 3.2015639656457204, | |
| "learning_rate": 4.0540540540540545e-05, | |
| "loss": 0.2704, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.23648648648648649, | |
| "grad_norm": 3.8305742015126354, | |
| "learning_rate": 4.72972972972973e-05, | |
| "loss": 0.2625, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2702702702702703, | |
| "grad_norm": 2.4187710692208926, | |
| "learning_rate": 4.9997978016429605e-05, | |
| "loss": 0.2675, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.30405405405405406, | |
| "grad_norm": 2.635487937809038, | |
| "learning_rate": 4.9985622766211935e-05, | |
| "loss": 0.2499, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.33783783783783783, | |
| "grad_norm": 2.316225114816826, | |
| "learning_rate": 4.996204175076325e-05, | |
| "loss": 0.2621, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3716216216216216, | |
| "grad_norm": 2.146594431614266, | |
| "learning_rate": 4.99272467427147e-05, | |
| "loss": 0.2884, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.40540540540540543, | |
| "grad_norm": 2.398859815430172, | |
| "learning_rate": 4.9881255113192526e-05, | |
| "loss": 0.2398, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4391891891891892, | |
| "grad_norm": 2.954151145902332, | |
| "learning_rate": 4.982408982314565e-05, | |
| "loss": 0.2543, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.47297297297297297, | |
| "grad_norm": 1.7929864319507005, | |
| "learning_rate": 4.975577941188258e-05, | |
| "loss": 0.2502, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5067567567567568, | |
| "grad_norm": 1.840568109556661, | |
| "learning_rate": 4.967635798282344e-05, | |
| "loss": 0.277, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5405405405405406, | |
| "grad_norm": 1.771348286216397, | |
| "learning_rate": 4.958586518647407e-05, | |
| "loss": 0.2518, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5743243243243243, | |
| "grad_norm": 2.623290643440281, | |
| "learning_rate": 4.9484346200630855e-05, | |
| "loss": 0.269, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.6081081081081081, | |
| "grad_norm": 2.6715602356902712, | |
| "learning_rate": 4.937185170782607e-05, | |
| "loss": 0.2469, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6418918918918919, | |
| "grad_norm": 1.5618566633614335, | |
| "learning_rate": 4.9248437870025035e-05, | |
| "loss": 0.2389, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6756756756756757, | |
| "grad_norm": 1.6043766275036224, | |
| "learning_rate": 4.911416630058772e-05, | |
| "loss": 0.2034, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7094594594594594, | |
| "grad_norm": 1.745799222586184, | |
| "learning_rate": 4.896910403350873e-05, | |
| "loss": 0.2718, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7432432432432432, | |
| "grad_norm": 1.6346813492093817, | |
| "learning_rate": 4.88133234899512e-05, | |
| "loss": 0.2441, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.777027027027027, | |
| "grad_norm": 2.0436191339698486, | |
| "learning_rate": 4.864690244209105e-05, | |
| "loss": 0.2135, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.8108108108108109, | |
| "grad_norm": 1.4009046058211432, | |
| "learning_rate": 4.8469923974289874e-05, | |
| "loss": 0.2855, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8445945945945946, | |
| "grad_norm": 2.1420433385829907, | |
| "learning_rate": 4.828247644161577e-05, | |
| "loss": 0.2443, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8783783783783784, | |
| "grad_norm": 2.029711117623661, | |
| "learning_rate": 4.808465342573274e-05, | |
| "loss": 0.2413, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9121621621621622, | |
| "grad_norm": 2.216487452899117, | |
| "learning_rate": 4.787655368818087e-05, | |
| "loss": 0.2136, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9459459459459459, | |
| "grad_norm": 1.9397316210901858, | |
| "learning_rate": 4.765828112107034e-05, | |
| "loss": 0.2541, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9797297297297297, | |
| "grad_norm": 1.2829673381620292, | |
| "learning_rate": 4.742994469521421e-05, | |
| "loss": 0.224, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.0135135135135136, | |
| "grad_norm": 1.8944274655064726, | |
| "learning_rate": 4.719165840572557e-05, | |
| "loss": 0.2169, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0472972972972974, | |
| "grad_norm": 1.7790710619051624, | |
| "learning_rate": 4.694354121510644e-05, | |
| "loss": 0.1452, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0810810810810811, | |
| "grad_norm": 1.5418464625119979, | |
| "learning_rate": 4.668571699385668e-05, | |
| "loss": 0.1531, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.114864864864865, | |
| "grad_norm": 1.5780966926016784, | |
| "learning_rate": 4.641831445863265e-05, | |
| "loss": 0.1694, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.1486486486486487, | |
| "grad_norm": 4.6428360428861515, | |
| "learning_rate": 4.614146710798645e-05, | |
| "loss": 0.1619, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.1824324324324325, | |
| "grad_norm": 1.4418340555320202, | |
| "learning_rate": 4.585531315571788e-05, | |
| "loss": 0.1611, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2162162162162162, | |
| "grad_norm": 2.0529270411153333, | |
| "learning_rate": 4.555999546187229e-05, | |
| "loss": 0.1383, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.1879187127096535, | |
| "learning_rate": 4.5255661461418854e-05, | |
| "loss": 0.1448, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2837837837837838, | |
| "grad_norm": 1.0143016739135113, | |
| "learning_rate": 4.4942463090644896e-05, | |
| "loss": 0.1465, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3175675675675675, | |
| "grad_norm": 2.090466049666836, | |
| "learning_rate": 4.462055671130289e-05, | |
| "loss": 0.1615, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3513513513513513, | |
| "grad_norm": 1.5608930671166168, | |
| "learning_rate": 4.4290103032548094e-05, | |
| "loss": 0.1412, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.385135135135135, | |
| "grad_norm": 1.2368408671304418, | |
| "learning_rate": 4.395126703070589e-05, | |
| "loss": 0.1571, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.4189189189189189, | |
| "grad_norm": 1.0805073107430125, | |
| "learning_rate": 4.360421786690862e-05, | |
| "loss": 0.1518, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.4527027027027026, | |
| "grad_norm": 1.9178479550142697, | |
| "learning_rate": 4.324912880264326e-05, | |
| "loss": 0.1476, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.4864864864864864, | |
| "grad_norm": 1.2390954008536137, | |
| "learning_rate": 4.288617711325207e-05, | |
| "loss": 0.1436, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.5202702702702702, | |
| "grad_norm": 1.2743622522980955, | |
| "learning_rate": 4.251554399942928e-05, | |
| "loss": 0.1494, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.554054054054054, | |
| "grad_norm": 1.1175039450703055, | |
| "learning_rate": 4.21374144967581e-05, | |
| "loss": 0.1482, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.5878378378378377, | |
| "grad_norm": 0.8990095650984934, | |
| "learning_rate": 4.1751977383333224e-05, | |
| "loss": 0.1412, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.6216216216216215, | |
| "grad_norm": 1.242740863487836, | |
| "learning_rate": 4.1359425085514906e-05, | |
| "loss": 0.1366, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.6554054054054053, | |
| "grad_norm": 1.1355786140547892, | |
| "learning_rate": 4.095995358186162e-05, | |
| "loss": 0.1667, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.689189189189189, | |
| "grad_norm": 1.1926029161081027, | |
| "learning_rate": 4.055376230528936e-05, | |
| "loss": 0.1308, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.722972972972973, | |
| "grad_norm": 1.0477428608330859, | |
| "learning_rate": 4.0141054043506406e-05, | |
| "loss": 0.1588, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.7567567567567568, | |
| "grad_norm": 2.434364212868205, | |
| "learning_rate": 3.972203483777315e-05, | |
| "loss": 0.1454, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.7905405405405406, | |
| "grad_norm": 1.025288482977718, | |
| "learning_rate": 3.929691388003772e-05, | |
| "loss": 0.1464, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.8243243243243243, | |
| "grad_norm": 1.0185166028149573, | |
| "learning_rate": 3.886590340849852e-05, | |
| "loss": 0.1228, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.8581081081081081, | |
| "grad_norm": 0.8097111792595093, | |
| "learning_rate": 3.842921860164607e-05, | |
| "loss": 0.13, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.8918918918918919, | |
| "grad_norm": 1.0056347069881086, | |
| "learning_rate": 3.798707747083694e-05, | |
| "loss": 0.1333, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.9256756756756757, | |
| "grad_norm": 1.0801697150500822, | |
| "learning_rate": 3.753970075145322e-05, | |
| "loss": 0.1323, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.9594594594594594, | |
| "grad_norm": 1.1000500869217074, | |
| "learning_rate": 3.7087311792702265e-05, | |
| "loss": 0.1337, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.9932432432432432, | |
| "grad_norm": 1.5122187810870562, | |
| "learning_rate": 3.663013644611139e-05, | |
| "loss": 0.1353, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.027027027027027, | |
| "grad_norm": 0.9578127795346132, | |
| "learning_rate": 3.616840295277328e-05, | |
| "loss": 0.0908, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.060810810810811, | |
| "grad_norm": 0.5934461784804044, | |
| "learning_rate": 3.5702341829398525e-05, | |
| "loss": 0.0851, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.0945945945945947, | |
| "grad_norm": 1.0725119181506966, | |
| "learning_rate": 3.523218575323198e-05, | |
| "loss": 0.0929, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.1283783783783785, | |
| "grad_norm": 0.7850220371401365, | |
| "learning_rate": 3.475816944589058e-05, | |
| "loss": 0.0911, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.1621621621621623, | |
| "grad_norm": 0.8527224202602178, | |
| "learning_rate": 3.4280529556180404e-05, | |
| "loss": 0.0929, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.195945945945946, | |
| "grad_norm": 1.04326596490438, | |
| "learning_rate": 3.379950454195172e-05, | |
| "loss": 0.0758, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.22972972972973, | |
| "grad_norm": 1.0501264893771993, | |
| "learning_rate": 3.331533455105084e-05, | |
| "loss": 0.0818, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.2635135135135136, | |
| "grad_norm": 0.8205870434886383, | |
| "learning_rate": 3.2828261301428206e-05, | |
| "loss": 0.0977, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.2972972972972974, | |
| "grad_norm": 0.8849982112270377, | |
| "learning_rate": 3.23385279604627e-05, | |
| "loss": 0.0719, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.331081081081081, | |
| "grad_norm": 0.7183927930700003, | |
| "learning_rate": 3.18463790235623e-05, | |
| "loss": 0.0873, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.364864864864865, | |
| "grad_norm": 0.7716121512809517, | |
| "learning_rate": 3.135206019210167e-05, | |
| "loss": 0.0771, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.3986486486486487, | |
| "grad_norm": 1.139603479425331, | |
| "learning_rate": 3.085581825075782e-05, | |
| "loss": 0.0709, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.4324324324324325, | |
| "grad_norm": 1.4786085416010715, | |
| "learning_rate": 3.0357900944304774e-05, | |
| "loss": 0.0902, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.4662162162162162, | |
| "grad_norm": 0.9071145610807975, | |
| "learning_rate": 2.9858556853929048e-05, | |
| "loss": 0.0833, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.7033676015416719, | |
| "learning_rate": 2.9358035273127483e-05, | |
| "loss": 0.0713, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.5337837837837838, | |
| "grad_norm": 0.673070630957069, | |
| "learning_rate": 2.8856586083249487e-05, | |
| "loss": 0.0672, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.5675675675675675, | |
| "grad_norm": 0.8537399411767074, | |
| "learning_rate": 2.83544596287458e-05, | |
| "loss": 0.0859, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.6013513513513513, | |
| "grad_norm": 0.5796181912861205, | |
| "learning_rate": 2.785190659218604e-05, | |
| "loss": 0.0794, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.635135135135135, | |
| "grad_norm": 0.6511959977506814, | |
| "learning_rate": 2.7349177869107462e-05, | |
| "loss": 0.078, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.668918918918919, | |
| "grad_norm": 0.975876535380539, | |
| "learning_rate": 2.684652444275741e-05, | |
| "loss": 0.0646, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.7027027027027026, | |
| "grad_norm": 0.6915954278323423, | |
| "learning_rate": 2.634419725879193e-05, | |
| "loss": 0.0753, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.7364864864864864, | |
| "grad_norm": 0.6789310321110773, | |
| "learning_rate": 2.58424470999932e-05, | |
| "loss": 0.0825, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.77027027027027, | |
| "grad_norm": 0.8180023876519836, | |
| "learning_rate": 2.534152446106825e-05, | |
| "loss": 0.0744, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.804054054054054, | |
| "grad_norm": 0.7821847539221951, | |
| "learning_rate": 2.4841679423591523e-05, | |
| "loss": 0.0684, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.8378378378378377, | |
| "grad_norm": 0.6251388492073695, | |
| "learning_rate": 2.4343161531153647e-05, | |
| "loss": 0.0764, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.8716216216216215, | |
| "grad_norm": 0.5954138849180285, | |
| "learning_rate": 2.3846219664778824e-05, | |
| "loss": 0.0739, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.9054054054054053, | |
| "grad_norm": 0.7069008525263534, | |
| "learning_rate": 2.3351101918672985e-05, | |
| "loss": 0.0772, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.939189189189189, | |
| "grad_norm": 0.6296050119251395, | |
| "learning_rate": 2.2858055476364822e-05, | |
| "loss": 0.0675, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.972972972972973, | |
| "grad_norm": 0.6716148252050718, | |
| "learning_rate": 2.2367326487301317e-05, | |
| "loss": 0.0612, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.0067567567567566, | |
| "grad_norm": 0.41293433850786937, | |
| "learning_rate": 2.1879159943959686e-05, | |
| "loss": 0.0583, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 3.0405405405405403, | |
| "grad_norm": 0.48242208510460555, | |
| "learning_rate": 2.139379955953686e-05, | |
| "loss": 0.04, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.074324324324324, | |
| "grad_norm": 0.5002995507342688, | |
| "learning_rate": 2.0911487646277623e-05, | |
| "loss": 0.0379, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 3.108108108108108, | |
| "grad_norm": 0.7307976680525754, | |
| "learning_rate": 2.0432464994502203e-05, | |
| "loss": 0.0402, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.141891891891892, | |
| "grad_norm": 0.4456729161007252, | |
| "learning_rate": 1.995697075239365e-05, | |
| "loss": 0.0417, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 3.175675675675676, | |
| "grad_norm": 0.5552051849040736, | |
| "learning_rate": 1.9485242306605028e-05, | |
| "loss": 0.0368, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.2094594594594597, | |
| "grad_norm": 0.48465471615450373, | |
| "learning_rate": 1.9017515163746058e-05, | |
| "loss": 0.0398, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.2432432432432434, | |
| "grad_norm": 0.40973628373713084, | |
| "learning_rate": 1.855402283280836e-05, | |
| "loss": 0.0395, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.277027027027027, | |
| "grad_norm": 0.5006185178881579, | |
| "learning_rate": 1.8094996708587958e-05, | |
| "loss": 0.0334, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 3.310810810810811, | |
| "grad_norm": 0.5022161933000103, | |
| "learning_rate": 1.7640665956163306e-05, | |
| "loss": 0.0331, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.3445945945945947, | |
| "grad_norm": 0.45753568314378656, | |
| "learning_rate": 1.719125739648648e-05, | |
| "loss": 0.0317, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 3.3783783783783785, | |
| "grad_norm": 0.448289241118566, | |
| "learning_rate": 1.6746995393144668e-05, | |
| "loss": 0.0315, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.4121621621621623, | |
| "grad_norm": 0.8295893160466691, | |
| "learning_rate": 1.6308101740348433e-05, | |
| "loss": 0.0386, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 3.445945945945946, | |
| "grad_norm": 0.5295192301457442, | |
| "learning_rate": 1.5874795552202773e-05, | |
| "loss": 0.0388, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 3.47972972972973, | |
| "grad_norm": 0.5769268856553119, | |
| "learning_rate": 1.5447293153316163e-05, | |
| "loss": 0.0318, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 3.5135135135135136, | |
| "grad_norm": 1.7958739927680984, | |
| "learning_rate": 1.5025807970802252e-05, | |
| "loss": 0.0335, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 3.5472972972972974, | |
| "grad_norm": 0.5185543173255863, | |
| "learning_rate": 1.4610550427728103e-05, | |
| "loss": 0.0323, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.581081081081081, | |
| "grad_norm": 0.6094438977040642, | |
| "learning_rate": 1.4201727838062181e-05, | |
| "loss": 0.0382, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 3.614864864864865, | |
| "grad_norm": 0.9983111528521604, | |
| "learning_rate": 1.3799544303174514e-05, | |
| "loss": 0.043, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 3.6486486486486487, | |
| "grad_norm": 0.5800713261881625, | |
| "learning_rate": 1.3404200609940754e-05, | |
| "loss": 0.0357, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 3.6824324324324325, | |
| "grad_norm": 0.3822107351668231, | |
| "learning_rate": 1.3015894130500977e-05, | |
| "loss": 0.033, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 3.7162162162162162, | |
| "grad_norm": 0.5208532366640735, | |
| "learning_rate": 1.2634818723723174e-05, | |
| "loss": 0.0336, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.6114967345181951, | |
| "learning_rate": 1.2261164638420832e-05, | |
| "loss": 0.0311, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 3.7837837837837838, | |
| "grad_norm": 0.5437365394783563, | |
| "learning_rate": 1.1895118418372734e-05, | |
| "loss": 0.0354, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 3.8175675675675675, | |
| "grad_norm": 0.3780362895703916, | |
| "learning_rate": 1.1536862809192518e-05, | |
| "loss": 0.0346, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 3.8513513513513513, | |
| "grad_norm": 3.518297058582475, | |
| "learning_rate": 1.1186576667094342e-05, | |
| "loss": 0.0285, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 3.885135135135135, | |
| "grad_norm": 0.3660586029820935, | |
| "learning_rate": 1.0844434869600428e-05, | |
| "loss": 0.0312, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.918918918918919, | |
| "grad_norm": 0.4588190949336144, | |
| "learning_rate": 1.0510608228234848e-05, | |
| "loss": 0.0327, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 3.9527027027027026, | |
| "grad_norm": 0.46633602361920473, | |
| "learning_rate": 1.0185263403247256e-05, | |
| "loss": 0.0344, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 3.9864864864864864, | |
| "grad_norm": 0.48024481445354344, | |
| "learning_rate": 9.868562820409103e-06, | |
| "loss": 0.0312, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.02027027027027, | |
| "grad_norm": 0.3049063524706929, | |
| "learning_rate": 9.560664589923895e-06, | |
| "loss": 0.0236, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 4.054054054054054, | |
| "grad_norm": 0.36714803546431, | |
| "learning_rate": 9.261722427491953e-06, | |
| "loss": 0.0149, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.087837837837838, | |
| "grad_norm": 0.2893986636191623, | |
| "learning_rate": 8.971885577569058e-06, | |
| "loss": 0.0125, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 4.121621621621622, | |
| "grad_norm": 0.2225194619419826, | |
| "learning_rate": 8.691298738857432e-06, | |
| "loss": 0.0105, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.155405405405405, | |
| "grad_norm": 0.271632458626091, | |
| "learning_rate": 8.420101992066028e-06, | |
| "loss": 0.0132, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 4.1891891891891895, | |
| "grad_norm": 0.38853806983200195, | |
| "learning_rate": 8.158430729976372e-06, | |
| "loss": 0.0117, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.222972972972973, | |
| "grad_norm": 0.28121579810924724, | |
| "learning_rate": 7.906415589848834e-06, | |
| "loss": 0.0119, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 4.256756756756757, | |
| "grad_norm": 0.36950427724679225, | |
| "learning_rate": 7.664182388203037e-06, | |
| "loss": 0.0122, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.29054054054054, | |
| "grad_norm": 0.40982200614975045, | |
| "learning_rate": 7.4318520580049444e-06, | |
| "loss": 0.0144, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 4.324324324324325, | |
| "grad_norm": 0.30667616284588944, | |
| "learning_rate": 7.209540588292083e-06, | |
| "loss": 0.0149, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 4.358108108108108, | |
| "grad_norm": 0.27670270061009083, | |
| "learning_rate": 6.9973589662669455e-06, | |
| "loss": 0.0097, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 4.391891891891892, | |
| "grad_norm": 1.156032111625232, | |
| "learning_rate": 6.7954131218875404e-06, | |
| "loss": 0.0141, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.425675675675675, | |
| "grad_norm": 0.43940200076280117, | |
| "learning_rate": 6.603803874982687e-06, | |
| "loss": 0.0187, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 4.45945945945946, | |
| "grad_norm": 0.5834566768734173, | |
| "learning_rate": 6.422626884918559e-06, | |
| "loss": 0.0117, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 4.493243243243243, | |
| "grad_norm": 0.38705922881740046, | |
| "learning_rate": 6.2519726028415145e-06, | |
| "loss": 0.0138, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 4.527027027027027, | |
| "grad_norm": 0.28730538588311555, | |
| "learning_rate": 6.091926226521089e-06, | |
| "loss": 0.0127, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 4.5608108108108105, | |
| "grad_norm": 0.3749162518326673, | |
| "learning_rate": 5.942567657815696e-06, | |
| "loss": 0.012, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 4.594594594594595, | |
| "grad_norm": 0.2566159254010731, | |
| "learning_rate": 5.8039714627822754e-06, | |
| "loss": 0.0089, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 4.628378378378378, | |
| "grad_norm": 0.4339033109119829, | |
| "learning_rate": 5.676206834449797e-06, | |
| "loss": 0.0113, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 4.662162162162162, | |
| "grad_norm": 0.3802169399499887, | |
| "learning_rate": 5.55933755827518e-06, | |
| "loss": 0.0098, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 4.695945945945946, | |
| "grad_norm": 0.4733511291475961, | |
| "learning_rate": 5.453421980298957e-06, | |
| "loss": 0.0105, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 4.72972972972973, | |
| "grad_norm": 0.3236015566148624, | |
| "learning_rate": 5.358512978016445e-06, | |
| "loss": 0.0096, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.763513513513513, | |
| "grad_norm": 0.47392413673934136, | |
| "learning_rate": 5.27465793397911e-06, | |
| "loss": 0.0125, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 4.797297297297297, | |
| "grad_norm": 0.24957782619954136, | |
| "learning_rate": 5.201898712139201e-06, | |
| "loss": 0.01, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 4.831081081081081, | |
| "grad_norm": 0.23348067557635, | |
| "learning_rate": 5.1402716369495194e-06, | |
| "loss": 0.0089, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 4.864864864864865, | |
| "grad_norm": 0.34646806534712726, | |
| "learning_rate": 5.089807475228711e-06, | |
| "loss": 0.0149, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 4.898648648648649, | |
| "grad_norm": 0.27931985147442506, | |
| "learning_rate": 5.050531420801205e-06, | |
| "loss": 0.0098, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 4.9324324324324325, | |
| "grad_norm": 0.38114606449587424, | |
| "learning_rate": 5.022463081919386e-06, | |
| "loss": 0.0141, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 4.966216216216216, | |
| "grad_norm": 0.3191715681389194, | |
| "learning_rate": 5.005616471474332e-06, | |
| "loss": 0.0106, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.29441671421186116, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0107, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 740, | |
| "total_flos": 435868890169344.0, | |
| "train_loss": 0.11167457006066232, | |
| "train_runtime": 12339.9122, | |
| "train_samples_per_second": 3.833, | |
| "train_steps_per_second": 0.06 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 740, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 435868890169344.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |