{ "best_metric": 0.7937743190661478, "best_model_checkpoint": "distilbert-base-multilingual-cased-hyper-matt/run-7naa0m57/checkpoint-1200", "epoch": 3.0, "eval_steps": 500, "global_step": 1200, "is_hyper_param_search": true, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.025, "grad_norm": 4.0042338371276855, "learning_rate": 4.843201071989088e-05, "loss": 0.6448, "step": 10 }, { "epoch": 0.05, "grad_norm": 9.521341323852539, "learning_rate": 4.8025019033169106e-05, "loss": 0.5938, "step": 20 }, { "epoch": 0.075, "grad_norm": 3.919041156768799, "learning_rate": 4.7618027346447335e-05, "loss": 0.6308, "step": 30 }, { "epoch": 0.1, "grad_norm": 3.4371910095214844, "learning_rate": 4.7211035659725565e-05, "loss": 0.5535, "step": 40 }, { "epoch": 0.125, "grad_norm": 4.366234302520752, "learning_rate": 4.6804043973003794e-05, "loss": 0.4918, "step": 50 }, { "epoch": 0.15, "grad_norm": 8.843981742858887, "learning_rate": 4.639705228628202e-05, "loss": 0.4615, "step": 60 }, { "epoch": 0.175, "grad_norm": 17.52216911315918, "learning_rate": 4.5990060599560246e-05, "loss": 0.2985, "step": 70 }, { "epoch": 0.2, "grad_norm": 0.17047187685966492, "learning_rate": 4.5583068912838476e-05, "loss": 0.4669, "step": 80 }, { "epoch": 0.225, "grad_norm": 8.667387008666992, "learning_rate": 4.5176077226116705e-05, "loss": 0.2397, "step": 90 }, { "epoch": 0.25, "grad_norm": 8.633122444152832, "learning_rate": 4.476908553939493e-05, "loss": 0.8068, "step": 100 }, { "epoch": 0.275, "grad_norm": 26.120525360107422, "learning_rate": 4.436209385267316e-05, "loss": 0.3711, "step": 110 }, { "epoch": 0.3, "grad_norm": 12.639058113098145, "learning_rate": 4.395510216595139e-05, "loss": 0.6239, "step": 120 }, { "epoch": 0.325, "grad_norm": 0.3218459188938141, "learning_rate": 4.3548110479229617e-05, "loss": 0.4321, "step": 130 }, { "epoch": 0.35, "grad_norm": 10.446372985839844, "learning_rate": 4.3141118792507846e-05, "loss": 0.6092, "step": 140 }, { "epoch": 0.375, "grad_norm": 2.5795059204101562, "learning_rate": 4.2734127105786075e-05, "loss": 0.5735, "step": 150 }, { "epoch": 0.4, "grad_norm": 1.5706645250320435, "learning_rate": 4.23271354190643e-05, "loss": 0.3974, "step": 160 }, { "epoch": 0.425, "grad_norm": 28.565317153930664, "learning_rate": 4.192014373234253e-05, "loss": 0.5174, "step": 170 }, { "epoch": 0.45, "grad_norm": 0.3528432548046112, "learning_rate": 4.151315204562076e-05, "loss": 0.3349, "step": 180 }, { "epoch": 0.475, "grad_norm": 0.09431620687246323, "learning_rate": 4.1106160358898987e-05, "loss": 0.522, "step": 190 }, { "epoch": 0.5, "grad_norm": 17.289031982421875, "learning_rate": 4.0699168672177216e-05, "loss": 0.4926, "step": 200 }, { "epoch": 0.525, "grad_norm": 4.251012802124023, "learning_rate": 4.029217698545544e-05, "loss": 0.3561, "step": 210 }, { "epoch": 0.55, "grad_norm": 7.21864652633667, "learning_rate": 3.988518529873367e-05, "loss": 0.4018, "step": 220 }, { "epoch": 0.575, "grad_norm": 18.046688079833984, "learning_rate": 3.94781936120119e-05, "loss": 0.4736, "step": 230 }, { "epoch": 0.6, "grad_norm": 8.16905403137207, "learning_rate": 3.907120192529013e-05, "loss": 0.5017, "step": 240 }, { "epoch": 0.625, "grad_norm": 26.532289505004883, "learning_rate": 3.866421023856835e-05, "loss": 0.3272, "step": 250 }, { "epoch": 0.65, "grad_norm": 42.285945892333984, "learning_rate": 3.825721855184658e-05, "loss": 0.3567, "step": 260 }, { "epoch": 0.675, "grad_norm": 10.441671371459961, "learning_rate": 3.785022686512481e-05, "loss": 0.8528, "step": 270 }, { "epoch": 0.7, "grad_norm": 10.992928504943848, "learning_rate": 3.744323517840304e-05, "loss": 0.5057, "step": 280 }, { "epoch": 0.725, "grad_norm": 0.5464246273040771, "learning_rate": 3.703624349168126e-05, "loss": 0.4161, "step": 290 }, { "epoch": 0.75, "grad_norm": 0.2611326277256012, "learning_rate": 3.662925180495949e-05, "loss": 0.0977, "step": 300 }, { "epoch": 0.775, "grad_norm": 0.4221794009208679, "learning_rate": 3.622226011823772e-05, "loss": 0.2858, "step": 310 }, { "epoch": 0.8, "grad_norm": 20.115922927856445, "learning_rate": 3.581526843151594e-05, "loss": 0.5096, "step": 320 }, { "epoch": 0.825, "grad_norm": 5.4764275550842285, "learning_rate": 3.540827674479417e-05, "loss": 0.5093, "step": 330 }, { "epoch": 0.85, "grad_norm": 0.22513504326343536, "learning_rate": 3.50012850580724e-05, "loss": 0.2849, "step": 340 }, { "epoch": 0.875, "grad_norm": 9.947456359863281, "learning_rate": 3.459429337135063e-05, "loss": 0.4448, "step": 350 }, { "epoch": 0.9, "grad_norm": 17.894901275634766, "learning_rate": 3.4187301684628854e-05, "loss": 1.1117, "step": 360 }, { "epoch": 0.925, "grad_norm": 0.4906502962112427, "learning_rate": 3.378030999790708e-05, "loss": 0.2753, "step": 370 }, { "epoch": 0.95, "grad_norm": 5.391515731811523, "learning_rate": 3.337331831118531e-05, "loss": 0.1935, "step": 380 }, { "epoch": 0.975, "grad_norm": 2.0084915161132812, "learning_rate": 3.296632662446354e-05, "loss": 0.3626, "step": 390 }, { "epoch": 1.0, "grad_norm": 12.514921188354492, "learning_rate": 3.2559334937741765e-05, "loss": 0.735, "step": 400 }, { "epoch": 1.0, "eval_accuracy": 0.7875, "eval_f1": 0.7301587301587301, "eval_loss": 0.7141464948654175, "eval_precision": 0.5989583333333334, "eval_recall": 0.9349593495934959, "eval_runtime": 1.5243, "eval_samples_per_second": 262.418, "eval_steps_per_second": 16.401, "step": 400 }, { "epoch": 1.025, "grad_norm": 11.676839828491211, "learning_rate": 3.2152343251019994e-05, "loss": 0.3937, "step": 410 }, { "epoch": 1.05, "grad_norm": 6.971975803375244, "learning_rate": 3.1745351564298224e-05, "loss": 0.4352, "step": 420 }, { "epoch": 1.075, "grad_norm": 0.2390863448381424, "learning_rate": 3.133835987757645e-05, "loss": 0.2793, "step": 430 }, { "epoch": 1.1, "grad_norm": 0.1350153088569641, "learning_rate": 3.093136819085468e-05, "loss": 0.5645, "step": 440 }, { "epoch": 1.125, "grad_norm": 0.14674435555934906, "learning_rate": 3.0524376504132905e-05, "loss": 0.0916, "step": 450 }, { "epoch": 1.15, "grad_norm": 0.0798887237906456, "learning_rate": 3.0117384817411138e-05, "loss": 0.1877, "step": 460 }, { "epoch": 1.175, "grad_norm": 0.7960764169692993, "learning_rate": 2.971039313068936e-05, "loss": 0.306, "step": 470 }, { "epoch": 1.2, "grad_norm": 0.27005836367607117, "learning_rate": 2.930340144396759e-05, "loss": 0.5818, "step": 480 }, { "epoch": 1.225, "grad_norm": 6.420020580291748, "learning_rate": 2.889640975724582e-05, "loss": 0.3856, "step": 490 }, { "epoch": 1.25, "grad_norm": 0.0868489220738411, "learning_rate": 2.848941807052405e-05, "loss": 0.4574, "step": 500 }, { "epoch": 1.275, "grad_norm": 0.12724563479423523, "learning_rate": 2.8082426383802275e-05, "loss": 0.5441, "step": 510 }, { "epoch": 1.3, "grad_norm": 16.045637130737305, "learning_rate": 2.7675434697080505e-05, "loss": 0.4005, "step": 520 }, { "epoch": 1.325, "grad_norm": 0.8872644901275635, "learning_rate": 2.726844301035873e-05, "loss": 0.1532, "step": 530 }, { "epoch": 1.35, "grad_norm": 0.1785900890827179, "learning_rate": 2.686145132363696e-05, "loss": 0.1264, "step": 540 }, { "epoch": 1.375, "grad_norm": 0.10678562521934509, "learning_rate": 2.6454459636915186e-05, "loss": 0.1142, "step": 550 }, { "epoch": 1.4, "grad_norm": 0.07380059361457825, "learning_rate": 2.6047467950193416e-05, "loss": 0.4046, "step": 560 }, { "epoch": 1.425, "grad_norm": 5.716458797454834, "learning_rate": 2.5640476263471645e-05, "loss": 0.0074, "step": 570 }, { "epoch": 1.45, "grad_norm": 64.4542007446289, "learning_rate": 2.5233484576749875e-05, "loss": 0.7207, "step": 580 }, { "epoch": 1.475, "grad_norm": 0.7955445647239685, "learning_rate": 2.4826492890028097e-05, "loss": 0.3275, "step": 590 }, { "epoch": 1.5, "grad_norm": 20.74042320251465, "learning_rate": 2.4419501203306327e-05, "loss": 0.4701, "step": 600 }, { "epoch": 1.525, "grad_norm": 0.36268430948257446, "learning_rate": 2.4012509516584553e-05, "loss": 0.1863, "step": 610 }, { "epoch": 1.55, "grad_norm": 0.19778229296207428, "learning_rate": 2.3605517829862782e-05, "loss": 0.2283, "step": 620 }, { "epoch": 1.575, "grad_norm": 0.08780460059642792, "learning_rate": 2.319852614314101e-05, "loss": 0.1166, "step": 630 }, { "epoch": 1.6, "grad_norm": 12.093305587768555, "learning_rate": 2.2791534456419238e-05, "loss": 0.1709, "step": 640 }, { "epoch": 1.625, "grad_norm": 5.422878265380859, "learning_rate": 2.2384542769697464e-05, "loss": 0.667, "step": 650 }, { "epoch": 1.65, "grad_norm": 13.302699089050293, "learning_rate": 2.1977551082975694e-05, "loss": 0.3448, "step": 660 }, { "epoch": 1.675, "grad_norm": 0.15557953715324402, "learning_rate": 2.1570559396253923e-05, "loss": 0.4754, "step": 670 }, { "epoch": 1.7, "grad_norm": 47.38977813720703, "learning_rate": 2.116356770953215e-05, "loss": 0.3435, "step": 680 }, { "epoch": 1.725, "grad_norm": 0.6711559891700745, "learning_rate": 2.075657602281038e-05, "loss": 0.1915, "step": 690 }, { "epoch": 1.75, "grad_norm": 49.03521728515625, "learning_rate": 2.0349584336088608e-05, "loss": 0.1532, "step": 700 }, { "epoch": 1.775, "grad_norm": 15.385904312133789, "learning_rate": 1.9942592649366834e-05, "loss": 0.6809, "step": 710 }, { "epoch": 1.8, "grad_norm": 30.673913955688477, "learning_rate": 1.9535600962645064e-05, "loss": 0.2397, "step": 720 }, { "epoch": 1.825, "grad_norm": 0.08131251484155655, "learning_rate": 1.912860927592329e-05, "loss": 0.3932, "step": 730 }, { "epoch": 1.85, "grad_norm": 16.99766731262207, "learning_rate": 1.872161758920152e-05, "loss": 0.2055, "step": 740 }, { "epoch": 1.875, "grad_norm": 0.846882700920105, "learning_rate": 1.8314625902479745e-05, "loss": 0.4532, "step": 750 }, { "epoch": 1.9, "grad_norm": 0.139489084482193, "learning_rate": 1.790763421575797e-05, "loss": 0.0558, "step": 760 }, { "epoch": 1.925, "grad_norm": 10.560323715209961, "learning_rate": 1.75006425290362e-05, "loss": 0.6155, "step": 770 }, { "epoch": 1.95, "grad_norm": 0.5508285760879517, "learning_rate": 1.7093650842314427e-05, "loss": 0.1549, "step": 780 }, { "epoch": 1.975, "grad_norm": 36.786476135253906, "learning_rate": 1.6686659155592656e-05, "loss": 0.4735, "step": 790 }, { "epoch": 2.0, "grad_norm": 0.08513189852237701, "learning_rate": 1.6279667468870882e-05, "loss": 0.0196, "step": 800 }, { "epoch": 2.0, "eval_accuracy": 0.8625, "eval_f1": 0.7859922178988327, "eval_loss": 0.48169881105422974, "eval_precision": 0.753731343283582, "eval_recall": 0.8211382113821138, "eval_runtime": 1.522, "eval_samples_per_second": 262.817, "eval_steps_per_second": 16.426, "step": 800 }, { "epoch": 2.025, "grad_norm": 2.163971424102783, "learning_rate": 1.5872675782149112e-05, "loss": 0.211, "step": 810 }, { "epoch": 2.05, "grad_norm": 3.05389404296875, "learning_rate": 1.546568409542734e-05, "loss": 0.2051, "step": 820 }, { "epoch": 2.075, "grad_norm": 66.39513397216797, "learning_rate": 1.5058692408705569e-05, "loss": 0.1411, "step": 830 }, { "epoch": 2.1, "grad_norm": 0.11405447870492935, "learning_rate": 1.4651700721983795e-05, "loss": 0.0306, "step": 840 }, { "epoch": 2.125, "grad_norm": 0.03801630809903145, "learning_rate": 1.4244709035262025e-05, "loss": 0.0271, "step": 850 }, { "epoch": 2.15, "grad_norm": 0.3133273422718048, "learning_rate": 1.3837717348540252e-05, "loss": 0.2901, "step": 860 }, { "epoch": 2.175, "grad_norm": 0.07565152645111084, "learning_rate": 1.343072566181848e-05, "loss": 0.2415, "step": 870 }, { "epoch": 2.2, "grad_norm": 84.18476867675781, "learning_rate": 1.3023733975096708e-05, "loss": 0.2952, "step": 880 }, { "epoch": 2.225, "grad_norm": 0.11311139911413193, "learning_rate": 1.2616742288374937e-05, "loss": 0.5811, "step": 890 }, { "epoch": 2.25, "grad_norm": 14.530881881713867, "learning_rate": 1.2209750601653163e-05, "loss": 0.3296, "step": 900 }, { "epoch": 2.275, "grad_norm": 24.885290145874023, "learning_rate": 1.1802758914931391e-05, "loss": 0.1924, "step": 910 }, { "epoch": 2.3, "grad_norm": 0.055552396923303604, "learning_rate": 1.1395767228209619e-05, "loss": 0.2252, "step": 920 }, { "epoch": 2.325, "grad_norm": 0.8402379751205444, "learning_rate": 1.0988775541487847e-05, "loss": 0.4053, "step": 930 }, { "epoch": 2.35, "grad_norm": 32.10145568847656, "learning_rate": 1.0581783854766075e-05, "loss": 0.1891, "step": 940 }, { "epoch": 2.375, "grad_norm": 45.39471435546875, "learning_rate": 1.0174792168044304e-05, "loss": 0.0353, "step": 950 }, { "epoch": 2.4, "grad_norm": 0.07263578474521637, "learning_rate": 9.767800481322532e-06, "loss": 0.1106, "step": 960 }, { "epoch": 2.425, "grad_norm": 0.032821133732795715, "learning_rate": 9.36080879460076e-06, "loss": 0.3095, "step": 970 }, { "epoch": 2.45, "grad_norm": 0.025584425777196884, "learning_rate": 8.953817107878986e-06, "loss": 0.1076, "step": 980 }, { "epoch": 2.475, "grad_norm": 0.0450764037668705, "learning_rate": 8.546825421157213e-06, "loss": 0.0046, "step": 990 }, { "epoch": 2.5, "grad_norm": 0.0470331646502018, "learning_rate": 8.139833734435441e-06, "loss": 0.1165, "step": 1000 }, { "epoch": 2.525, "grad_norm": 0.04697234183549881, "learning_rate": 7.73284204771367e-06, "loss": 0.0749, "step": 1010 }, { "epoch": 2.55, "grad_norm": 0.04398488625884056, "learning_rate": 7.3258503609918976e-06, "loss": 0.2088, "step": 1020 }, { "epoch": 2.575, "grad_norm": 0.046139203011989594, "learning_rate": 6.918858674270126e-06, "loss": 0.206, "step": 1030 }, { "epoch": 2.6, "grad_norm": 0.0412064827978611, "learning_rate": 6.511866987548354e-06, "loss": 0.3262, "step": 1040 }, { "epoch": 2.625, "grad_norm": 13.24852466583252, "learning_rate": 6.104875300826582e-06, "loss": 0.3108, "step": 1050 }, { "epoch": 2.65, "grad_norm": 0.028375335037708282, "learning_rate": 5.6978836141048095e-06, "loss": 0.0038, "step": 1060 }, { "epoch": 2.675, "grad_norm": 0.202249214053154, "learning_rate": 5.290891927383037e-06, "loss": 0.1516, "step": 1070 }, { "epoch": 2.7, "grad_norm": 0.02920701913535595, "learning_rate": 4.883900240661266e-06, "loss": 0.1316, "step": 1080 }, { "epoch": 2.725, "grad_norm": 0.08516351878643036, "learning_rate": 4.476908553939493e-06, "loss": 0.0915, "step": 1090 }, { "epoch": 2.75, "grad_norm": 0.03471991792321205, "learning_rate": 4.069916867217721e-06, "loss": 0.0177, "step": 1100 }, { "epoch": 2.775, "grad_norm": 0.06554219126701355, "learning_rate": 3.6629251804959488e-06, "loss": 0.2439, "step": 1110 }, { "epoch": 2.8, "grad_norm": 0.2342367321252823, "learning_rate": 3.255933493774177e-06, "loss": 0.1958, "step": 1120 }, { "epoch": 2.825, "grad_norm": 0.02388150431215763, "learning_rate": 2.8489418070524047e-06, "loss": 0.0264, "step": 1130 }, { "epoch": 2.85, "grad_norm": 0.10175933688879013, "learning_rate": 2.441950120330633e-06, "loss": 0.2652, "step": 1140 }, { "epoch": 2.875, "grad_norm": 50.01603317260742, "learning_rate": 2.0349584336088603e-06, "loss": 0.3156, "step": 1150 }, { "epoch": 2.9, "grad_norm": 0.07987692207098007, "learning_rate": 1.6279667468870885e-06, "loss": 0.0203, "step": 1160 }, { "epoch": 2.925, "grad_norm": 20.490392684936523, "learning_rate": 1.2209750601653165e-06, "loss": 0.1305, "step": 1170 }, { "epoch": 2.95, "grad_norm": 2.933145046234131, "learning_rate": 8.139833734435442e-07, "loss": 0.0051, "step": 1180 }, { "epoch": 2.975, "grad_norm": 0.023663857951760292, "learning_rate": 4.069916867217721e-07, "loss": 0.0157, "step": 1190 }, { "epoch": 3.0, "grad_norm": 1.89351487159729, "learning_rate": 0.0, "loss": 0.0026, "step": 1200 }, { "epoch": 3.0, "eval_accuracy": 0.8675, "eval_f1": 0.7937743190661478, "eval_loss": 0.5679348111152649, "eval_precision": 0.7611940298507462, "eval_recall": 0.8292682926829268, "eval_runtime": 1.5586, "eval_samples_per_second": 256.645, "eval_steps_per_second": 16.04, "step": 1200 } ], "logging_steps": 10, "max_steps": 1200, "num_input_tokens_seen": 0, "num_train_epochs": 3, "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": 635446111352832.0, "train_batch_size": 4, "trial_name": null, "trial_params": { "_wandb": {}, "assignments": {}, "learning_rate": 4.8839002406612654e-05, "metric": "eval/loss", "num_train_epochs": 3, "per_device_train_batch_size": 4, "seed": 1 } }