| { | |
| "best_metric": 0.22245945036411285, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-swagen-baseline-model/checkpoint-2200", | |
| "epoch": 6.205250596658711, | |
| "eval_steps": 100, | |
| "global_step": 2600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2386634844868735, | |
| "grad_norm": 19.838756561279297, | |
| "learning_rate": 0.00028799999999999995, | |
| "loss": 15.3971, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2386634844868735, | |
| "eval_loss": 3.594007968902588, | |
| "eval_runtime": 32.6156, | |
| "eval_samples_per_second": 17.446, | |
| "eval_steps_per_second": 4.384, | |
| "eval_wer": 1.0050711917300565, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.477326968973747, | |
| "grad_norm": 6.821050643920898, | |
| "learning_rate": 0.00029769045709703284, | |
| "loss": 6.4342, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.477326968973747, | |
| "eval_loss": 2.992433547973633, | |
| "eval_runtime": 32.5813, | |
| "eval_samples_per_second": 17.464, | |
| "eval_steps_per_second": 4.389, | |
| "eval_wer": 0.9869319290033158, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7159904534606205, | |
| "grad_norm": 5.072504997253418, | |
| "learning_rate": 0.00029528468323977545, | |
| "loss": 3.4197, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7159904534606205, | |
| "eval_loss": 0.2737194001674652, | |
| "eval_runtime": 32.4531, | |
| "eval_samples_per_second": 17.533, | |
| "eval_steps_per_second": 4.406, | |
| "eval_wer": 0.20226253169494832, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.954653937947494, | |
| "grad_norm": 1.5491737127304077, | |
| "learning_rate": 0.000292878909382518, | |
| "loss": 0.56, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.954653937947494, | |
| "eval_loss": 0.2543294429779053, | |
| "eval_runtime": 32.2994, | |
| "eval_samples_per_second": 17.616, | |
| "eval_steps_per_second": 4.427, | |
| "eval_wer": 0.1962161107860347, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.1933174224343674, | |
| "grad_norm": 2.1193718910217285, | |
| "learning_rate": 0.0002904731355252606, | |
| "loss": 0.5187, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.1933174224343674, | |
| "eval_loss": 0.2419968843460083, | |
| "eval_runtime": 32.686, | |
| "eval_samples_per_second": 17.408, | |
| "eval_steps_per_second": 4.375, | |
| "eval_wer": 0.1929003315779208, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.431980906921241, | |
| "grad_norm": 1.6395008563995361, | |
| "learning_rate": 0.0002880673616680032, | |
| "loss": 0.5115, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.431980906921241, | |
| "eval_loss": 0.2393239438533783, | |
| "eval_runtime": 32.6608, | |
| "eval_samples_per_second": 17.421, | |
| "eval_steps_per_second": 4.378, | |
| "eval_wer": 0.19465574409986347, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.6706443914081146, | |
| "grad_norm": 1.2931002378463745, | |
| "learning_rate": 0.00028566158781074576, | |
| "loss": 0.5086, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.6706443914081146, | |
| "eval_loss": 0.23598293960094452, | |
| "eval_runtime": 32.3728, | |
| "eval_samples_per_second": 17.576, | |
| "eval_steps_per_second": 4.417, | |
| "eval_wer": 0.1891944606982641, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.9093078758949882, | |
| "grad_norm": 1.0657984018325806, | |
| "learning_rate": 0.0002832558139534883, | |
| "loss": 0.4801, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.9093078758949882, | |
| "eval_loss": 0.2332880049943924, | |
| "eval_runtime": 32.406, | |
| "eval_samples_per_second": 17.558, | |
| "eval_steps_per_second": 4.413, | |
| "eval_wer": 0.18743904817632143, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.1479713603818618, | |
| "grad_norm": 0.8465839624404907, | |
| "learning_rate": 0.0002808740978348035, | |
| "loss": 0.5281, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.1479713603818618, | |
| "eval_loss": 0.2355046272277832, | |
| "eval_runtime": 32.703, | |
| "eval_samples_per_second": 17.399, | |
| "eval_steps_per_second": 4.373, | |
| "eval_wer": 0.1958260191144919, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.386634844868735, | |
| "grad_norm": 1.4161683320999146, | |
| "learning_rate": 0.00027849238171611865, | |
| "loss": 0.4683, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.386634844868735, | |
| "eval_loss": 0.23779602348804474, | |
| "eval_runtime": 32.6391, | |
| "eval_samples_per_second": 17.433, | |
| "eval_steps_per_second": 4.381, | |
| "eval_wer": 0.1956309732787205, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.6252983293556085, | |
| "grad_norm": 0.9228078126907349, | |
| "learning_rate": 0.0002760866078588612, | |
| "loss": 0.4548, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.6252983293556085, | |
| "eval_loss": 0.22832728922367096, | |
| "eval_runtime": 32.4057, | |
| "eval_samples_per_second": 17.559, | |
| "eval_steps_per_second": 4.413, | |
| "eval_wer": 0.18743904817632143, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.863961813842482, | |
| "grad_norm": 1.0473681688308716, | |
| "learning_rate": 0.00027368083400160383, | |
| "loss": 0.4654, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.863961813842482, | |
| "eval_loss": 0.23233628273010254, | |
| "eval_runtime": 32.4947, | |
| "eval_samples_per_second": 17.511, | |
| "eval_steps_per_second": 4.401, | |
| "eval_wer": 0.1891944606982641, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.1026252983293556, | |
| "grad_norm": 1.9320465326309204, | |
| "learning_rate": 0.0002712750601443464, | |
| "loss": 0.453, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.1026252983293556, | |
| "eval_loss": 0.22875255346298218, | |
| "eval_runtime": 32.7978, | |
| "eval_samples_per_second": 17.349, | |
| "eval_steps_per_second": 4.36, | |
| "eval_wer": 0.1897795982055783, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.341288782816229, | |
| "grad_norm": 3.4785170555114746, | |
| "learning_rate": 0.000268869286287089, | |
| "loss": 0.4542, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.341288782816229, | |
| "eval_loss": 0.2302933633327484, | |
| "eval_runtime": 32.7775, | |
| "eval_samples_per_second": 17.359, | |
| "eval_steps_per_second": 4.363, | |
| "eval_wer": 0.19016968987712113, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.579952267303103, | |
| "grad_norm": 2.0417723655700684, | |
| "learning_rate": 0.00026646351242983157, | |
| "loss": 0.4621, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.579952267303103, | |
| "eval_loss": 0.22530591487884521, | |
| "eval_runtime": 32.5076, | |
| "eval_samples_per_second": 17.504, | |
| "eval_steps_per_second": 4.399, | |
| "eval_wer": 0.1864638189974644, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.8186157517899764, | |
| "grad_norm": 1.299824833869934, | |
| "learning_rate": 0.00026405773857257413, | |
| "loss": 0.4342, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.8186157517899764, | |
| "eval_loss": 0.22665663063526154, | |
| "eval_runtime": 32.7877, | |
| "eval_samples_per_second": 17.354, | |
| "eval_steps_per_second": 4.361, | |
| "eval_wer": 0.18685391066900722, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.05727923627685, | |
| "grad_norm": 1.9400004148483276, | |
| "learning_rate": 0.00026165196471531675, | |
| "loss": 0.466, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.05727923627685, | |
| "eval_loss": 0.2283923178911209, | |
| "eval_runtime": 32.8561, | |
| "eval_samples_per_second": 17.318, | |
| "eval_steps_per_second": 4.352, | |
| "eval_wer": 0.1897795982055783, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.2959427207637235, | |
| "grad_norm": 1.079958200454712, | |
| "learning_rate": 0.0002592461908580593, | |
| "loss": 0.4268, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.2959427207637235, | |
| "eval_loss": 0.23249927163124084, | |
| "eval_runtime": 32.7252, | |
| "eval_samples_per_second": 17.387, | |
| "eval_steps_per_second": 4.37, | |
| "eval_wer": 0.1958260191144919, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.534606205250597, | |
| "grad_norm": 3.2974212169647217, | |
| "learning_rate": 0.0002568404170008019, | |
| "loss": 0.4283, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.534606205250597, | |
| "eval_loss": 0.22498926520347595, | |
| "eval_runtime": 32.5783, | |
| "eval_samples_per_second": 17.466, | |
| "eval_steps_per_second": 4.389, | |
| "eval_wer": 0.18860932319094986, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.77326968973747, | |
| "grad_norm": 3.1521952152252197, | |
| "learning_rate": 0.0002544346431435445, | |
| "loss": 0.4407, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.77326968973747, | |
| "eval_loss": 0.2249845415353775, | |
| "eval_runtime": 32.4737, | |
| "eval_samples_per_second": 17.522, | |
| "eval_steps_per_second": 4.404, | |
| "eval_wer": 0.18841427735517846, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.011933174224343, | |
| "grad_norm": 1.2635365724563599, | |
| "learning_rate": 0.00025202886928628706, | |
| "loss": 0.4762, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.011933174224343, | |
| "eval_loss": 0.22774527966976166, | |
| "eval_runtime": 33.4042, | |
| "eval_samples_per_second": 17.034, | |
| "eval_steps_per_second": 4.281, | |
| "eval_wer": 0.1893895065340355, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.250596658711217, | |
| "grad_norm": 1.1228593587875366, | |
| "learning_rate": 0.0002496230954290297, | |
| "loss": 0.4289, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.250596658711217, | |
| "eval_loss": 0.22245945036411285, | |
| "eval_runtime": 32.7331, | |
| "eval_samples_per_second": 17.383, | |
| "eval_steps_per_second": 4.369, | |
| "eval_wer": 0.18724400234055003, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.4892601431980905, | |
| "grad_norm": 0.8405218124389648, | |
| "learning_rate": 0.00024721732157177224, | |
| "loss": 0.4391, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.4892601431980905, | |
| "eval_loss": 0.22291946411132812, | |
| "eval_runtime": 32.5165, | |
| "eval_samples_per_second": 17.499, | |
| "eval_steps_per_second": 4.398, | |
| "eval_wer": 0.18841427735517846, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.727923627684964, | |
| "grad_norm": 4.196171283721924, | |
| "learning_rate": 0.0002448115477145148, | |
| "loss": 0.4333, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 5.727923627684964, | |
| "eval_loss": 0.22290098667144775, | |
| "eval_runtime": 32.4851, | |
| "eval_samples_per_second": 17.516, | |
| "eval_steps_per_second": 4.402, | |
| "eval_wer": 0.18782913984786426, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 5.966587112171838, | |
| "grad_norm": 0.8724733591079712, | |
| "learning_rate": 0.0002424057738572574, | |
| "loss": 0.4351, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.966587112171838, | |
| "eval_loss": 0.22788961231708527, | |
| "eval_runtime": 32.9988, | |
| "eval_samples_per_second": 17.243, | |
| "eval_steps_per_second": 4.333, | |
| "eval_wer": 0.19016968987712113, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.205250596658711, | |
| "grad_norm": 3.7559807300567627, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 0.4065, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.205250596658711, | |
| "eval_loss": 0.22799032926559448, | |
| "eval_runtime": 33.0675, | |
| "eval_samples_per_second": 17.207, | |
| "eval_steps_per_second": 4.324, | |
| "eval_wer": 0.19387556075677784, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.205250596658711, | |
| "step": 2600, | |
| "total_flos": 1.068815361623146e+19, | |
| "train_loss": 1.3819761775090145, | |
| "train_runtime": 3531.6962, | |
| "train_samples_per_second": 28.448, | |
| "train_steps_per_second": 3.559 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 12570, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 4, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 2 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.068815361623146e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |