| { | |
| "best_metric": 92.12121212121212, | |
| "best_model_checkpoint": "./whisper-bangla-dialect/checkpoint-1504", | |
| "epoch": 7.978779840848806, | |
| "eval_steps": 500, | |
| "global_step": 1504, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.26525198938992045, | |
| "grad_norm": 28.63650894165039, | |
| "learning_rate": 2.35e-06, | |
| "loss": 4.9503, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5305039787798409, | |
| "grad_norm": 20.406896591186523, | |
| "learning_rate": 4.85e-06, | |
| "loss": 2.9269, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7957559681697612, | |
| "grad_norm": 7.10680627822876, | |
| "learning_rate": 7.350000000000001e-06, | |
| "loss": 2.3479, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9973474801061007, | |
| "eval_loss": 1.5939799547195435, | |
| "eval_norm_levenshtein_similarity": 83.03030303030303, | |
| "eval_runtime": 219.3411, | |
| "eval_samples_per_second": 1.527, | |
| "eval_steps_per_second": 0.383, | |
| "eval_wer": 26.519337016574585, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.0610079575596818, | |
| "grad_norm": 6.035281181335449, | |
| "learning_rate": 9.85e-06, | |
| "loss": 2.0388, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3262599469496021, | |
| "grad_norm": 5.856254577636719, | |
| "learning_rate": 1.2350000000000002e-05, | |
| "loss": 1.843, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.5915119363395225, | |
| "grad_norm": 5.164959907531738, | |
| "learning_rate": 1.4850000000000002e-05, | |
| "loss": 1.7357, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.8567639257294428, | |
| "grad_norm": 5.301006317138672, | |
| "learning_rate": 1.735e-05, | |
| "loss": 1.6723, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.5232762098312378, | |
| "eval_norm_levenshtein_similarity": 87.22222222222223, | |
| "eval_runtime": 218.1243, | |
| "eval_samples_per_second": 1.536, | |
| "eval_steps_per_second": 0.385, | |
| "eval_wer": 18.95027624309392, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.1220159151193636, | |
| "grad_norm": 4.089552402496338, | |
| "learning_rate": 1.985e-05, | |
| "loss": 1.6156, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.387267904509284, | |
| "grad_norm": 3.2860209941864014, | |
| "learning_rate": 1.991069416392988e-05, | |
| "loss": 1.5792, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.6525198938992043, | |
| "grad_norm": 3.775080442428589, | |
| "learning_rate": 1.9621456437066593e-05, | |
| "loss": 1.5522, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.9177718832891246, | |
| "grad_norm": 2.990183115005493, | |
| "learning_rate": 1.913776795338998e-05, | |
| "loss": 1.538, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.9973474801061006, | |
| "eval_loss": 1.5003355741500854, | |
| "eval_norm_levenshtein_similarity": 87.15151515151516, | |
| "eval_runtime": 219.6769, | |
| "eval_samples_per_second": 1.525, | |
| "eval_steps_per_second": 0.382, | |
| "eval_wer": 18.011049723756905, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 3.183023872679045, | |
| "grad_norm": 2.5845985412597656, | |
| "learning_rate": 1.8469404113571235e-05, | |
| "loss": 1.5167, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.4482758620689653, | |
| "grad_norm": 2.177119731903076, | |
| "learning_rate": 1.7629872628769222e-05, | |
| "loss": 1.5024, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.713527851458886, | |
| "grad_norm": 2.521852970123291, | |
| "learning_rate": 1.6636140528212427e-05, | |
| "loss": 1.4945, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.9787798408488064, | |
| "grad_norm": 2.6489243507385254, | |
| "learning_rate": 1.550829125355928e-05, | |
| "loss": 1.4874, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.4792444705963135, | |
| "eval_norm_levenshtein_similarity": 90.77777777777779, | |
| "eval_runtime": 217.1116, | |
| "eval_samples_per_second": 1.543, | |
| "eval_steps_per_second": 0.387, | |
| "eval_wer": 12.430939226519337, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 4.244031830238727, | |
| "grad_norm": 1.9068670272827148, | |
| "learning_rate": 1.4269118770199764e-05, | |
| "loss": 1.4761, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.5092838196286475, | |
| "grad_norm": 1.6706056594848633, | |
| "learning_rate": 1.294366689855624e-05, | |
| "loss": 1.4734, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 4.774535809018568, | |
| "grad_norm": 1.9710417985916138, | |
| "learning_rate": 1.155872317555169e-05, | |
| "loss": 1.4665, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.997347480106101, | |
| "eval_loss": 1.4712462425231934, | |
| "eval_norm_levenshtein_similarity": 91.06060606060606, | |
| "eval_runtime": 217.2231, | |
| "eval_samples_per_second": 1.542, | |
| "eval_steps_per_second": 0.387, | |
| "eval_wer": 12.486187845303867, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 5.039787798408488, | |
| "grad_norm": 2.245420217514038, | |
| "learning_rate": 1.0142277475364053e-05, | |
| "loss": 1.4638, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 5.305039787798409, | |
| "grad_norm": 1.3094449043273926, | |
| "learning_rate": 8.722956330804456e-06, | |
| "loss": 1.4569, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.570291777188329, | |
| "grad_norm": 2.1748969554901123, | |
| "learning_rate": 7.32944438775016e-06, | |
| "loss": 1.4555, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 5.835543766578249, | |
| "grad_norm": 1.970129132270813, | |
| "learning_rate": 5.989904685105696e-06, | |
| "loss": 1.4518, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.4650499820709229, | |
| "eval_norm_levenshtein_similarity": 91.30303030303031, | |
| "eval_runtime": 217.2177, | |
| "eval_samples_per_second": 1.542, | |
| "eval_steps_per_second": 0.387, | |
| "eval_wer": 11.767955801104971, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 6.10079575596817, | |
| "grad_norm": 1.6859333515167236, | |
| "learning_rate": 4.7314094765017325e-06, | |
| "loss": 1.4496, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 6.36604774535809, | |
| "grad_norm": 1.515045166015625, | |
| "learning_rate": 3.579393096891569e-06, | |
| "loss": 1.4462, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 6.63129973474801, | |
| "grad_norm": 0.8732560276985168, | |
| "learning_rate": 2.5571379316751565e-06, | |
| "loss": 1.4454, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 6.896551724137931, | |
| "grad_norm": 1.4642410278320312, | |
| "learning_rate": 1.6853038769745466e-06, | |
| "loss": 1.4448, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 6.997347480106101, | |
| "eval_loss": 1.4623610973358154, | |
| "eval_norm_levenshtein_similarity": 91.96969696969697, | |
| "eval_runtime": 219.1609, | |
| "eval_samples_per_second": 1.529, | |
| "eval_steps_per_second": 0.383, | |
| "eval_wer": 10.607734806629834, | |
| "step": 1319 | |
| }, | |
| { | |
| "epoch": 7.161803713527852, | |
| "grad_norm": 1.0192619562149048, | |
| "learning_rate": 9.815108007235663e-07, | |
| "loss": 1.4432, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 7.427055702917772, | |
| "grad_norm": 1.5917143821716309, | |
| "learning_rate": 4.5998244308173126e-07, | |
| "loss": 1.4421, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 1.224084734916687, | |
| "learning_rate": 1.3125895298661705e-07, | |
| "loss": 1.4421, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 7.957559681697613, | |
| "grad_norm": 1.5394171476364136, | |
| "learning_rate": 1.9838705144537006e-09, | |
| "loss": 1.4419, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.978779840848806, | |
| "eval_loss": 1.462101697921753, | |
| "eval_norm_levenshtein_similarity": 92.12121212121212, | |
| "eval_runtime": 215.9027, | |
| "eval_samples_per_second": 1.552, | |
| "eval_steps_per_second": 0.389, | |
| "eval_wer": 10.441988950276244, | |
| "step": 1504 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 1504, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "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": 6.94249897549824e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |