| { | |
| "best_metric": 0.8064516129032258, | |
| "best_model_checkpoint": "distilbert-base-multilingual-cased-hyper-matt/run-rtpz8b71/checkpoint-800", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 800, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.5273020267486572, | |
| "learning_rate": 3.379454045530049e-05, | |
| "loss": 0.672, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 7.657852649688721, | |
| "learning_rate": 3.336676146219542e-05, | |
| "loss": 0.53, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.539841651916504, | |
| "learning_rate": 3.293898246909035e-05, | |
| "loss": 0.5481, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 3.097090244293213, | |
| "learning_rate": 3.251120347598528e-05, | |
| "loss": 0.2841, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 4.80979061126709, | |
| "learning_rate": 3.208342448288021e-05, | |
| "loss": 0.6117, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 7.640440464019775, | |
| "learning_rate": 3.1655645489775144e-05, | |
| "loss": 0.5832, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 3.1154446601867676, | |
| "learning_rate": 3.1227866496670074e-05, | |
| "loss": 0.4628, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.84663724899292, | |
| "learning_rate": 3.0800087503565005e-05, | |
| "loss": 0.4354, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 15.556452751159668, | |
| "learning_rate": 3.0372308510459932e-05, | |
| "loss": 0.3762, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 5.357182025909424, | |
| "learning_rate": 2.9944529517354863e-05, | |
| "loss": 0.4135, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 5.593770980834961, | |
| "learning_rate": 2.9516750524249797e-05, | |
| "loss": 0.3815, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 4.29806661605835, | |
| "learning_rate": 2.9088971531144725e-05, | |
| "loss": 0.4898, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.596214771270752, | |
| "learning_rate": 2.8661192538039655e-05, | |
| "loss": 0.2377, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 13.040666580200195, | |
| "learning_rate": 2.8233413544934586e-05, | |
| "loss": 0.4523, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.686233997344971, | |
| "learning_rate": 2.7805634551829517e-05, | |
| "loss": 0.2944, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.4252870082855225, | |
| "learning_rate": 2.7377855558724448e-05, | |
| "loss": 0.4242, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 5.249920845031738, | |
| "learning_rate": 2.695007656561938e-05, | |
| "loss": 0.3134, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.831853866577148, | |
| "learning_rate": 2.652229757251431e-05, | |
| "loss": 0.4533, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.20128059387207, | |
| "learning_rate": 2.6094518579409237e-05, | |
| "loss": 0.3014, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.756709098815918, | |
| "learning_rate": 2.5666739586304167e-05, | |
| "loss": 0.4128, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.845, | |
| "eval_f1": 0.6804123711340206, | |
| "eval_loss": 0.3612925708293915, | |
| "eval_precision": 0.9295774647887324, | |
| "eval_recall": 0.5365853658536586, | |
| "eval_runtime": 1.5276, | |
| "eval_samples_per_second": 261.844, | |
| "eval_steps_per_second": 16.365, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 5.490586757659912, | |
| "learning_rate": 2.52389605931991e-05, | |
| "loss": 0.3643, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 6.718981742858887, | |
| "learning_rate": 2.481118160009403e-05, | |
| "loss": 0.2173, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.35438114404678345, | |
| "learning_rate": 2.438340260698896e-05, | |
| "loss": 0.2357, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 7.001437664031982, | |
| "learning_rate": 2.395562361388389e-05, | |
| "loss": 0.2616, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 6.03153133392334, | |
| "learning_rate": 2.352784462077882e-05, | |
| "loss": 0.1749, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 29.80844497680664, | |
| "learning_rate": 2.3100065627673752e-05, | |
| "loss": 0.3992, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 3.6485345363616943, | |
| "learning_rate": 2.2672286634568683e-05, | |
| "loss": 0.1931, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.2925957143306732, | |
| "learning_rate": 2.2244507641463614e-05, | |
| "loss": 0.3383, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 11.773463249206543, | |
| "learning_rate": 2.181672864835854e-05, | |
| "loss": 0.4417, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.6896932125091553, | |
| "learning_rate": 2.1388949655253475e-05, | |
| "loss": 0.3196, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.8070700168609619, | |
| "learning_rate": 2.0961170662148406e-05, | |
| "loss": 0.1319, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.24381496012210846, | |
| "learning_rate": 2.0533391669043333e-05, | |
| "loss": 0.3669, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 7.7754669189453125, | |
| "learning_rate": 2.0105612675938267e-05, | |
| "loss": 0.397, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 3.648085594177246, | |
| "learning_rate": 1.9677833682833195e-05, | |
| "loss": 0.4326, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 19.058910369873047, | |
| "learning_rate": 1.9250054689728126e-05, | |
| "loss": 0.3497, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.9193096160888672, | |
| "learning_rate": 1.882227569662306e-05, | |
| "loss": 0.3383, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 5.150022029876709, | |
| "learning_rate": 1.8394496703517987e-05, | |
| "loss": 0.2809, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.5875343680381775, | |
| "learning_rate": 1.7966717710412918e-05, | |
| "loss": 0.2199, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 9.385665893554688, | |
| "learning_rate": 1.753893871730785e-05, | |
| "loss": 0.4043, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.588594913482666, | |
| "learning_rate": 1.711115972420278e-05, | |
| "loss": 0.1863, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.865, | |
| "eval_f1": 0.7906976744186046, | |
| "eval_loss": 0.33189040422439575, | |
| "eval_precision": 0.7555555555555555, | |
| "eval_recall": 0.8292682926829268, | |
| "eval_runtime": 1.5225, | |
| "eval_samples_per_second": 262.722, | |
| "eval_steps_per_second": 16.42, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.32351887226104736, | |
| "learning_rate": 1.668338073109771e-05, | |
| "loss": 0.1385, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.6892914175987244, | |
| "learning_rate": 1.625560173799264e-05, | |
| "loss": 0.3113, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 0.4579280912876129, | |
| "learning_rate": 1.5827822744887572e-05, | |
| "loss": 0.1769, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 9.545488357543945, | |
| "learning_rate": 1.5400043751782503e-05, | |
| "loss": 0.2754, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 2.9151611328125, | |
| "learning_rate": 1.4972264758677432e-05, | |
| "loss": 0.2701, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.5319097638130188, | |
| "learning_rate": 1.4544485765572362e-05, | |
| "loss": 0.098, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.9106752276420593, | |
| "learning_rate": 1.4116706772467293e-05, | |
| "loss": 0.0424, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 4.542623043060303, | |
| "learning_rate": 1.3688927779362224e-05, | |
| "loss": 0.3205, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 2.6717588901519775, | |
| "learning_rate": 1.3261148786257155e-05, | |
| "loss": 0.3126, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 63.507198333740234, | |
| "learning_rate": 1.2833369793152084e-05, | |
| "loss": 0.2623, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 33.924095153808594, | |
| "learning_rate": 1.2405590800047015e-05, | |
| "loss": 0.1062, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 8.15170669555664, | |
| "learning_rate": 1.1977811806941945e-05, | |
| "loss": 0.3471, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 34.99510192871094, | |
| "learning_rate": 1.1550032813836876e-05, | |
| "loss": 0.3089, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 9.204129219055176, | |
| "learning_rate": 1.1122253820731807e-05, | |
| "loss": 0.2824, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 7.578718185424805, | |
| "learning_rate": 1.0694474827626738e-05, | |
| "loss": 0.0619, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 3.9372365474700928, | |
| "learning_rate": 1.0266695834521667e-05, | |
| "loss": 0.341, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 5.813101291656494, | |
| "learning_rate": 9.838916841416597e-06, | |
| "loss": 0.0413, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 54.38895034790039, | |
| "learning_rate": 9.41113784831153e-06, | |
| "loss": 0.2031, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 42.406883239746094, | |
| "learning_rate": 8.983358855206459e-06, | |
| "loss": 0.2095, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.2973996102809906, | |
| "learning_rate": 8.55557986210139e-06, | |
| "loss": 0.0812, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8725, | |
| "eval_f1": 0.7866108786610879, | |
| "eval_loss": 0.4548227787017822, | |
| "eval_precision": 0.8103448275862069, | |
| "eval_recall": 0.7642276422764228, | |
| "eval_runtime": 1.5153, | |
| "eval_samples_per_second": 263.982, | |
| "eval_steps_per_second": 16.499, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.10320937633514404, | |
| "learning_rate": 8.12780086899632e-06, | |
| "loss": 0.0107, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.14922451972961426, | |
| "learning_rate": 7.700021875891251e-06, | |
| "loss": 0.0992, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 58.235931396484375, | |
| "learning_rate": 7.272242882786181e-06, | |
| "loss": 0.0839, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.5583060383796692, | |
| "learning_rate": 6.844463889681112e-06, | |
| "loss": 0.188, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.6185976266860962, | |
| "learning_rate": 6.416684896576042e-06, | |
| "loss": 0.1084, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.10940929502248764, | |
| "learning_rate": 5.988905903470973e-06, | |
| "loss": 0.0349, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 1.7489935159683228, | |
| "learning_rate": 5.561126910365903e-06, | |
| "loss": 0.0542, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.22723452746868134, | |
| "learning_rate": 5.133347917260833e-06, | |
| "loss": 0.1929, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.11978468298912048, | |
| "learning_rate": 4.705568924155765e-06, | |
| "loss": 0.0968, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.12295297533273697, | |
| "learning_rate": 4.277789931050695e-06, | |
| "loss": 0.0746, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 1.1583808660507202, | |
| "learning_rate": 3.850010937945626e-06, | |
| "loss": 0.0119, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.05860567465424538, | |
| "learning_rate": 3.422231944840556e-06, | |
| "loss": 0.1999, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.3195663094520569, | |
| "learning_rate": 2.9944529517354863e-06, | |
| "loss": 0.1388, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.07110361754894257, | |
| "learning_rate": 2.5666739586304167e-06, | |
| "loss": 0.2383, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.10332443565130234, | |
| "learning_rate": 2.1388949655253474e-06, | |
| "loss": 0.1156, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.222213014960289, | |
| "learning_rate": 1.711115972420278e-06, | |
| "loss": 0.0304, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 0.5618990659713745, | |
| "learning_rate": 1.2833369793152083e-06, | |
| "loss": 0.1119, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 5.86262845993042, | |
| "learning_rate": 8.55557986210139e-07, | |
| "loss": 0.1783, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 0.07000931352376938, | |
| "learning_rate": 4.277789931050695e-07, | |
| "loss": 0.0181, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.0716305747628212, | |
| "learning_rate": 0.0, | |
| "loss": 0.1302, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.88, | |
| "eval_f1": 0.8064516129032258, | |
| "eval_loss": 0.5255711078643799, | |
| "eval_precision": 0.8, | |
| "eval_recall": 0.8130081300813008, | |
| "eval_runtime": 3.5354, | |
| "eval_samples_per_second": 113.14, | |
| "eval_steps_per_second": 7.071, | |
| "step": 800 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 800, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 847261481803776.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": { | |
| "_wandb": {}, | |
| "assignments": {}, | |
| "learning_rate": 3.422231944840556e-05, | |
| "metric": "eval/loss", | |
| "num_train_epochs": 4, | |
| "per_device_train_batch_size": 8, | |
| "seed": 28 | |
| } | |
| } | |