| { | |
| "best_global_step": 234872, | |
| "best_metric": 0.3474995039635673, | |
| "best_model_checkpoint": "facebook-mms-1b-all-common_voice_fleurs-rw-200hrs-v1/checkpoint-234872", | |
| "epoch": 26.0, | |
| "eval_steps": 500, | |
| "global_step": 277576, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 8.114748001098633, | |
| "learning_rate": 9.993443237167478e-06, | |
| "loss": 1.7863, | |
| "step": 10676 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 0.11142080162244088, | |
| "eval_loss": 0.4298412799835205, | |
| "eval_runtime": 370.3842, | |
| "eval_samples_per_second": 36.206, | |
| "eval_steps_per_second": 2.265, | |
| "eval_wer": 0.38987518778522096, | |
| "step": 10676 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 49.95594787597656, | |
| "learning_rate": 1.9989696515548896e-05, | |
| "loss": 0.2748, | |
| "step": 21352 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 0.11646517201831753, | |
| "eval_loss": 0.42387428879737854, | |
| "eval_runtime": 330.7981, | |
| "eval_samples_per_second": 40.538, | |
| "eval_steps_per_second": 2.536, | |
| "eval_wer": 0.4003628152193426, | |
| "step": 21352 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 12.085474967956543, | |
| "learning_rate": 2.998594979393031e-05, | |
| "loss": 0.2545, | |
| "step": 32028 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 0.1220423125205042, | |
| "eval_loss": 0.4541724622249603, | |
| "eval_runtime": 351.2843, | |
| "eval_samples_per_second": 38.174, | |
| "eval_steps_per_second": 2.388, | |
| "eval_wer": 0.4101890607432043, | |
| "step": 32028 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 8.780159950256348, | |
| "learning_rate": 3.998313975271638e-05, | |
| "loss": 0.2486, | |
| "step": 42704 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.12309158494361855, | |
| "eval_loss": 0.460758239030838, | |
| "eval_runtime": 312.2958, | |
| "eval_samples_per_second": 42.94, | |
| "eval_steps_per_second": 2.687, | |
| "eval_wer": 0.4010241971296025, | |
| "step": 42704 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 4.321907043457031, | |
| "learning_rate": 4.997939303109779e-05, | |
| "loss": 0.2513, | |
| "step": 53380 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.1371239090413663, | |
| "eval_loss": 0.4873207211494446, | |
| "eval_runtime": 302.0004, | |
| "eval_samples_per_second": 44.404, | |
| "eval_steps_per_second": 2.778, | |
| "eval_wer": 0.4268936781337692, | |
| "step": 53380 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3.6101784706115723, | |
| "learning_rate": 4.947501528268029e-05, | |
| "loss": 0.2481, | |
| "step": 64056 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 0.1327763422959816, | |
| "eval_loss": 0.4789027273654938, | |
| "eval_runtime": 328.5795, | |
| "eval_samples_per_second": 40.812, | |
| "eval_steps_per_second": 2.553, | |
| "eval_wer": 0.41674618996778123, | |
| "step": 64056 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 4.900432109832764, | |
| "learning_rate": 4.89488473901126e-05, | |
| "loss": 0.2243, | |
| "step": 74732 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 0.12506541007030397, | |
| "eval_loss": 0.4427885413169861, | |
| "eval_runtime": 373.1444, | |
| "eval_samples_per_second": 35.938, | |
| "eval_steps_per_second": 2.248, | |
| "eval_wer": 0.3932765804665577, | |
| "step": 74732 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.762911558151245, | |
| "learning_rate": 4.842272879651358e-05, | |
| "loss": 0.2074, | |
| "step": 85408 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.1276913024263408, | |
| "eval_loss": 0.4647955298423767, | |
| "eval_runtime": 419.1893, | |
| "eval_samples_per_second": 31.99, | |
| "eval_steps_per_second": 2.001, | |
| "eval_wer": 0.39266244012131635, | |
| "step": 85408 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 4.190975666046143, | |
| "learning_rate": 4.789665950188322e-05, | |
| "loss": 0.1951, | |
| "step": 96084 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 0.12099710704476624, | |
| "eval_loss": 0.4297826886177063, | |
| "eval_runtime": 367.3195, | |
| "eval_samples_per_second": 36.508, | |
| "eval_steps_per_second": 2.284, | |
| "eval_wer": 0.3886941486597568, | |
| "step": 96084 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 12.867461204528809, | |
| "learning_rate": 4.7370491609315534e-05, | |
| "loss": 0.183, | |
| "step": 106760 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 0.12186065553769111, | |
| "eval_loss": 0.4488038122653961, | |
| "eval_runtime": 318.5142, | |
| "eval_samples_per_second": 42.102, | |
| "eval_steps_per_second": 2.634, | |
| "eval_wer": 0.3807292207976266, | |
| "step": 106760 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 2.9338295459747314, | |
| "learning_rate": 4.6844373015716514e-05, | |
| "loss": 0.1704, | |
| "step": 117436 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 0.11717146521268779, | |
| "eval_loss": 0.4291200637817383, | |
| "eval_runtime": 309.0823, | |
| "eval_samples_per_second": 43.386, | |
| "eval_steps_per_second": 2.714, | |
| "eval_wer": 0.36931565868914107, | |
| "step": 117436 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.193000316619873, | |
| "learning_rate": 4.631815582418016e-05, | |
| "loss": 0.1602, | |
| "step": 128112 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 0.11806890493374943, | |
| "eval_loss": 0.4337969422340393, | |
| "eval_runtime": 370.6877, | |
| "eval_samples_per_second": 36.176, | |
| "eval_steps_per_second": 2.263, | |
| "eval_wer": 0.3649127448294107, | |
| "step": 128112 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 4.6738057136535645, | |
| "learning_rate": 4.5792086529549806e-05, | |
| "loss": 0.1503, | |
| "step": 138788 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 0.11863963321557261, | |
| "eval_loss": 0.43087342381477356, | |
| "eval_runtime": 445.2671, | |
| "eval_samples_per_second": 30.117, | |
| "eval_steps_per_second": 1.884, | |
| "eval_wer": 0.36557412673967066, | |
| "step": 138788 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 4.588621616363525, | |
| "learning_rate": 4.526601723491945e-05, | |
| "loss": 0.1417, | |
| "step": 149464 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 0.11495497889254312, | |
| "eval_loss": 0.4210534691810608, | |
| "eval_runtime": 380.9506, | |
| "eval_samples_per_second": 35.201, | |
| "eval_steps_per_second": 2.202, | |
| "eval_wer": 0.36186093972921135, | |
| "step": 149464 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 20.85131072998047, | |
| "learning_rate": 4.473989864132043e-05, | |
| "loss": 0.136, | |
| "step": 160140 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 0.1191358007954949, | |
| "eval_loss": 0.44502612948417664, | |
| "eval_runtime": 358.3686, | |
| "eval_samples_per_second": 37.42, | |
| "eval_steps_per_second": 2.341, | |
| "eval_wer": 0.37130925273292453, | |
| "step": 160140 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.3467574417591095, | |
| "learning_rate": 4.421373074875274e-05, | |
| "loss": 0.1305, | |
| "step": 170816 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 0.11509596640159207, | |
| "eval_loss": 0.4436805248260498, | |
| "eval_runtime": 352.8374, | |
| "eval_samples_per_second": 38.006, | |
| "eval_steps_per_second": 2.378, | |
| "eval_wer": 0.36488439989039956, | |
| "step": 170816 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 2.450319290161133, | |
| "learning_rate": 4.368761215515372e-05, | |
| "loss": 0.122, | |
| "step": 181492 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_cer": 0.11497666927855064, | |
| "eval_loss": 0.45026466250419617, | |
| "eval_runtime": 361.6111, | |
| "eval_samples_per_second": 37.084, | |
| "eval_steps_per_second": 2.32, | |
| "eval_wer": 0.3567210574551914, | |
| "step": 181492 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 1.8293219804763794, | |
| "learning_rate": 4.316149356155469e-05, | |
| "loss": 0.1139, | |
| "step": 192168 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_cer": 0.11221792330821767, | |
| "eval_loss": 0.4319141209125519, | |
| "eval_runtime": 319.6496, | |
| "eval_samples_per_second": 41.952, | |
| "eval_steps_per_second": 2.625, | |
| "eval_wer": 0.3517795897542494, | |
| "step": 192168 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 121.26933288574219, | |
| "learning_rate": 4.263532566898701e-05, | |
| "loss": 0.1057, | |
| "step": 202844 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_cer": 0.11463504569893201, | |
| "eval_loss": 0.43899470567703247, | |
| "eval_runtime": 318.2195, | |
| "eval_samples_per_second": 42.141, | |
| "eval_steps_per_second": 2.637, | |
| "eval_wer": 0.35932879184421623, | |
| "step": 202844 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 2.178046703338623, | |
| "learning_rate": 4.2109157776419316e-05, | |
| "loss": 0.0997, | |
| "step": 213520 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_cer": 0.11560975742014548, | |
| "eval_loss": 0.45046666264533997, | |
| "eval_runtime": 309.141, | |
| "eval_samples_per_second": 43.378, | |
| "eval_steps_per_second": 2.714, | |
| "eval_wer": 0.35749581912149586, | |
| "step": 213520 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 82.85478973388672, | |
| "learning_rate": 4.158313778075763e-05, | |
| "loss": 0.0931, | |
| "step": 224196 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_cer": 0.11554197496387195, | |
| "eval_loss": 0.47222137451171875, | |
| "eval_runtime": 297.0885, | |
| "eval_samples_per_second": 45.138, | |
| "eval_steps_per_second": 2.824, | |
| "eval_wer": 0.35703285178431393, | |
| "step": 224196 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 4.658969402313232, | |
| "learning_rate": 4.105696988818994e-05, | |
| "loss": 0.0872, | |
| "step": 234872 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_cer": 0.1130882500467699, | |
| "eval_loss": 0.47128280997276306, | |
| "eval_runtime": 297.5307, | |
| "eval_samples_per_second": 45.071, | |
| "eval_steps_per_second": 2.82, | |
| "eval_wer": 0.3474995039635673, | |
| "step": 234872 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 31.182987213134766, | |
| "learning_rate": 4.053080199562225e-05, | |
| "loss": 0.0834, | |
| "step": 245548 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_cer": 0.11390977341680517, | |
| "eval_loss": 0.4635840058326721, | |
| "eval_runtime": 294.749, | |
| "eval_samples_per_second": 45.496, | |
| "eval_steps_per_second": 2.846, | |
| "eval_wer": 0.3590169975150937, | |
| "step": 245548 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 3.176016330718994, | |
| "learning_rate": 4.000478199996056e-05, | |
| "loss": 0.0815, | |
| "step": 256224 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_cer": 0.11560975742014548, | |
| "eval_loss": 0.46226194500923157, | |
| "eval_runtime": 299.7969, | |
| "eval_samples_per_second": 44.73, | |
| "eval_steps_per_second": 2.799, | |
| "eval_wer": 0.3642419146061471, | |
| "step": 256224 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 28.0598087310791, | |
| "learning_rate": 3.947851550945554e-05, | |
| "loss": 0.0764, | |
| "step": 266900 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_cer": 0.11135844176266922, | |
| "eval_loss": 0.45431631803512573, | |
| "eval_runtime": 289.8433, | |
| "eval_samples_per_second": 46.266, | |
| "eval_steps_per_second": 2.895, | |
| "eval_wer": 0.348793922845076, | |
| "step": 266900 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 5.333448886871338, | |
| "learning_rate": 3.895239691585652e-05, | |
| "loss": 0.0729, | |
| "step": 277576 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_cer": 0.11139911123643334, | |
| "eval_loss": 0.4733089804649353, | |
| "eval_runtime": 299.9429, | |
| "eval_samples_per_second": 44.709, | |
| "eval_steps_per_second": 2.797, | |
| "eval_wer": 0.34857661164599063, | |
| "step": 277576 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1067600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 10, | |
| "early_stopping_threshold": 0.001 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 4 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.748850144039913e+21, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |