| { | |
| "best_metric": 0.7969924812030075, | |
| "best_model_checkpoint": "distilbert-base-multilingual-cased-hyper-matt/run-os2whcgy/checkpoint-400", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.763803005218506, | |
| "learning_rate": 4.100144208506588e-05, | |
| "loss": 0.6566, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.878039836883545, | |
| "learning_rate": 4.058728610440865e-05, | |
| "loss": 0.5424, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.597256660461426, | |
| "learning_rate": 4.017313012375141e-05, | |
| "loss": 0.5403, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 3.657524824142456, | |
| "learning_rate": 3.975897414309418e-05, | |
| "loss": 0.2928, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 4.957123279571533, | |
| "learning_rate": 3.934481816243695e-05, | |
| "loss": 0.6895, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 3.636430501937866, | |
| "learning_rate": 3.893066218177972e-05, | |
| "loss": 0.5498, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 3.0615527629852295, | |
| "learning_rate": 3.851650620112249e-05, | |
| "loss": 0.4264, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.246954917907715, | |
| "learning_rate": 3.810235022046526e-05, | |
| "loss": 0.3793, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.777850866317749, | |
| "learning_rate": 3.768819423980803e-05, | |
| "loss": 0.3108, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 7.651516914367676, | |
| "learning_rate": 3.72740382591508e-05, | |
| "loss": 0.2877, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 6.894057750701904, | |
| "learning_rate": 3.6859882278493564e-05, | |
| "loss": 0.405, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.9369120597839355, | |
| "learning_rate": 3.6445726297836334e-05, | |
| "loss": 0.5737, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.7391217350959778, | |
| "learning_rate": 3.6031570317179104e-05, | |
| "loss": 0.2456, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 5.605640411376953, | |
| "learning_rate": 3.5617414336521874e-05, | |
| "loss": 0.511, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 7.593440532684326, | |
| "learning_rate": 3.520325835586464e-05, | |
| "loss": 0.3234, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.2000792026519775, | |
| "learning_rate": 3.478910237520741e-05, | |
| "loss": 0.4554, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 5.821305751800537, | |
| "learning_rate": 3.437494639455018e-05, | |
| "loss": 0.3331, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.795863628387451, | |
| "learning_rate": 3.396079041389295e-05, | |
| "loss": 0.4863, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 6.364365100860596, | |
| "learning_rate": 3.354663443323572e-05, | |
| "loss": 0.2883, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.494615077972412, | |
| "learning_rate": 3.313247845257849e-05, | |
| "loss": 0.4124, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.82, | |
| "eval_f1": 0.6, | |
| "eval_loss": 0.37046054005622864, | |
| "eval_precision": 0.9473684210526315, | |
| "eval_recall": 0.43902439024390244, | |
| "eval_runtime": 1.5233, | |
| "eval_samples_per_second": 262.59, | |
| "eval_steps_per_second": 16.412, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 2.7175021171569824, | |
| "learning_rate": 3.271832247192126e-05, | |
| "loss": 0.3539, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 8.024584770202637, | |
| "learning_rate": 3.230416649126403e-05, | |
| "loss": 0.257, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.2626192271709442, | |
| "learning_rate": 3.189001051060679e-05, | |
| "loss": 0.2057, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.23563814163208, | |
| "learning_rate": 3.147585452994956e-05, | |
| "loss": 0.1401, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 11.776407241821289, | |
| "learning_rate": 3.106169854929233e-05, | |
| "loss": 0.1677, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 12.863431930541992, | |
| "learning_rate": 3.06475425686351e-05, | |
| "loss": 0.3453, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.2926412522792816, | |
| "learning_rate": 3.0233386587977867e-05, | |
| "loss": 0.1243, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.21822890639305115, | |
| "learning_rate": 2.9819230607320637e-05, | |
| "loss": 0.2649, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 12.913641929626465, | |
| "learning_rate": 2.9405074626663404e-05, | |
| "loss": 0.4294, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.5034979581832886, | |
| "learning_rate": 2.8990918646006174e-05, | |
| "loss": 0.3872, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 13.819730758666992, | |
| "learning_rate": 2.857676266534894e-05, | |
| "loss": 0.1721, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.3381294012069702, | |
| "learning_rate": 2.8162606684691714e-05, | |
| "loss": 0.2673, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 10.247021675109863, | |
| "learning_rate": 2.7748450704034484e-05, | |
| "loss": 0.3316, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 4.321371078491211, | |
| "learning_rate": 2.733429472337725e-05, | |
| "loss": 0.3636, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 12.241943359375, | |
| "learning_rate": 2.692013874272002e-05, | |
| "loss": 0.3765, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.8232184052467346, | |
| "learning_rate": 2.650598276206279e-05, | |
| "loss": 0.3517, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 3.5735747814178467, | |
| "learning_rate": 2.6091826781405557e-05, | |
| "loss": 0.2621, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.5765199065208435, | |
| "learning_rate": 2.5677670800748327e-05, | |
| "loss": 0.2027, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 7.612463474273682, | |
| "learning_rate": 2.5263514820091094e-05, | |
| "loss": 0.3536, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.938665866851807, | |
| "learning_rate": 2.4849358839433864e-05, | |
| "loss": 0.175, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.865, | |
| "eval_f1": 0.7969924812030075, | |
| "eval_loss": 0.34953439235687256, | |
| "eval_precision": 0.7412587412587412, | |
| "eval_recall": 0.8617886178861789, | |
| "eval_runtime": 1.5177, | |
| "eval_samples_per_second": 263.56, | |
| "eval_steps_per_second": 16.473, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1000, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 423630740901888.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": { | |
| "_wandb": {}, | |
| "assignments": {}, | |
| "learning_rate": 4.141559806572311e-05, | |
| "metric": "eval/loss", | |
| "num_train_epochs": 5, | |
| "per_device_train_batch_size": 8, | |
| "seed": 28 | |
| } | |
| } | |