| { | |
| "best_metric": 0.7435250282287598, | |
| "best_model_checkpoint": "./whisper-small-arabic-finetuned-on-halabi_daataset_with-diacritics-2/checkpoint-200", | |
| "epoch": 17.548672566371682, | |
| "eval_steps": 200, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4424778761061947, | |
| "grad_norm": 49.0963249206543, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.9627, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.8849557522123894, | |
| "grad_norm": 8.113597869873047, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.44, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.3185840707964602, | |
| "grad_norm": 5.070877552032471, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.1815, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.7610619469026547, | |
| "grad_norm": 4.050365447998047, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1147, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.1946902654867255, | |
| "grad_norm": 6.347820281982422, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.0813, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.6371681415929205, | |
| "grad_norm": 4.465636730194092, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0812, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.0707964601769913, | |
| "grad_norm": 4.107247352600098, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.0708, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.5132743362831858, | |
| "grad_norm": 5.341292381286621, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0504, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.5132743362831858, | |
| "eval_loss": 0.7435250282287598, | |
| "eval_runtime": 94.7482, | |
| "eval_samples_per_second": 1.045, | |
| "eval_steps_per_second": 0.074, | |
| "eval_wer": 0.7221022318214543, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.9557522123893807, | |
| "grad_norm": 2.8596858978271484, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.0502, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.389380530973451, | |
| "grad_norm": 1.7310813665390015, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.0419, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.831858407079646, | |
| "grad_norm": 2.756486415863037, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.0396, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.265486725663717, | |
| "grad_norm": 6.585121154785156, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.0258, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.707964601769912, | |
| "grad_norm": 0.31891903281211853, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.0298, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.1415929203539825, | |
| "grad_norm": 1.1073646545410156, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.0226, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 6.584070796460177, | |
| "grad_norm": 1.3673633337020874, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.02, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 7.017699115044247, | |
| "grad_norm": 2.9860990047454834, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0138, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.017699115044247, | |
| "eval_loss": 0.9074220061302185, | |
| "eval_runtime": 84.7371, | |
| "eval_samples_per_second": 1.168, | |
| "eval_steps_per_second": 0.083, | |
| "eval_wer": 0.7134629229661628, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.460176991150442, | |
| "grad_norm": 0.045507967472076416, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.0095, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 7.902654867256637, | |
| "grad_norm": 0.6703155636787415, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.0128, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 8.336283185840708, | |
| "grad_norm": 0.6974866390228271, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.0097, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 8.778761061946902, | |
| "grad_norm": 2.5476107597351074, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.0083, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 9.212389380530974, | |
| "grad_norm": 0.13264885544776917, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.0053, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 9.654867256637168, | |
| "grad_norm": 0.18639236688613892, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.0073, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 10.08849557522124, | |
| "grad_norm": 0.33259156346321106, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.0067, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 10.530973451327434, | |
| "grad_norm": 2.248627185821533, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0049, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.530973451327434, | |
| "eval_loss": 1.182573676109314, | |
| "eval_runtime": 99.0092, | |
| "eval_samples_per_second": 1.0, | |
| "eval_steps_per_second": 0.071, | |
| "eval_wer": 0.7156227501799856, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.973451327433628, | |
| "grad_norm": 0.3167344629764557, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.0038, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 11.4070796460177, | |
| "grad_norm": 0.15390612185001373, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.0041, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 11.849557522123893, | |
| "grad_norm": 0.10117767751216888, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.0049, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 12.283185840707965, | |
| "grad_norm": 3.8471896648406982, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.0057, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 12.725663716814159, | |
| "grad_norm": 0.028711393475532532, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.0034, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 13.15929203539823, | |
| "grad_norm": 0.08290126919746399, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.0033, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 13.601769911504425, | |
| "grad_norm": 0.03170332685112953, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.0022, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 14.035398230088495, | |
| "grad_norm": 0.2151724398136139, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0013, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 14.035398230088495, | |
| "eval_loss": 1.1965606212615967, | |
| "eval_runtime": 88.1337, | |
| "eval_samples_per_second": 1.123, | |
| "eval_steps_per_second": 0.079, | |
| "eval_wer": 0.7156227501799856, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 14.47787610619469, | |
| "grad_norm": 0.03010549768805504, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.0007, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 14.920353982300885, | |
| "grad_norm": 0.02499246783554554, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.0015, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 15.353982300884956, | |
| "grad_norm": 0.017421068623661995, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.0009, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 15.79646017699115, | |
| "grad_norm": 0.14170809090137482, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.001, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 16.23008849557522, | |
| "grad_norm": 0.013867147266864777, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 0.0006, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 16.672566371681416, | |
| "grad_norm": 0.020731482654809952, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.0008, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 17.106194690265486, | |
| "grad_norm": 0.03430557996034622, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.0009, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 17.548672566371682, | |
| "grad_norm": 0.10500048846006393, | |
| "learning_rate": 0.0, | |
| "loss": 0.0008, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.548672566371682, | |
| "eval_loss": 1.1946815252304077, | |
| "eval_runtime": 98.7155, | |
| "eval_samples_per_second": 1.003, | |
| "eval_steps_per_second": 0.071, | |
| "eval_wer": 0.7163426925845933, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.548672566371682, | |
| "step": 1000, | |
| "total_flos": 9.15623758135296e+18, | |
| "train_loss": 0.05816659236606211, | |
| "train_runtime": 17237.6477, | |
| "train_samples_per_second": 1.856, | |
| "train_steps_per_second": 0.058 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 1000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 18, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.15623758135296e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |