| { | |
| "best_metric": 0.20998027920722961, | |
| "best_model_checkpoint": "/scratch/skscla001/results/mms-zeroshot-300m-genbed-f-model/checkpoint-5600", | |
| "epoch": 16.986301369863014, | |
| "eval_steps": 200, | |
| "global_step": 6200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.547945205479452, | |
| "eval_loss": 2.3236358165740967, | |
| "eval_runtime": 58.3404, | |
| "eval_samples_per_second": 16.627, | |
| "eval_steps_per_second": 2.091, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "eval_loss": 0.3330934941768646, | |
| "eval_runtime": 57.5798, | |
| "eval_samples_per_second": 16.846, | |
| "eval_steps_per_second": 2.119, | |
| "eval_wer": 0.5504350628424106, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 0.44414466619491577, | |
| "learning_rate": 0.00028902304147465435, | |
| "loss": 2.6731, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "eval_loss": 0.29691705107688904, | |
| "eval_runtime": 57.9342, | |
| "eval_samples_per_second": 16.743, | |
| "eval_steps_per_second": 2.106, | |
| "eval_wer": 0.5189601460951767, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "eval_loss": 0.2805863320827484, | |
| "eval_runtime": 57.8326, | |
| "eval_samples_per_second": 16.773, | |
| "eval_steps_per_second": 2.11, | |
| "eval_wer": 0.5121925018799012, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 0.2989409565925598, | |
| "learning_rate": 0.00027519815668202764, | |
| "loss": 0.4193, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "eval_loss": 0.2701254189014435, | |
| "eval_runtime": 57.9519, | |
| "eval_samples_per_second": 16.738, | |
| "eval_steps_per_second": 2.105, | |
| "eval_wer": 0.4741647867654958, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.287671232876712, | |
| "eval_loss": 0.27029839158058167, | |
| "eval_runtime": 57.8498, | |
| "eval_samples_per_second": 16.768, | |
| "eval_steps_per_second": 2.109, | |
| "eval_wer": 0.47695778279084755, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.8356164383561646, | |
| "eval_loss": 0.2574402093887329, | |
| "eval_runtime": 57.9028, | |
| "eval_samples_per_second": 16.752, | |
| "eval_steps_per_second": 2.107, | |
| "eval_wer": 0.47577613062627566, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.109589041095891, | |
| "grad_norm": 0.5375602841377258, | |
| "learning_rate": 0.00026137327188940093, | |
| "loss": 0.367, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.383561643835616, | |
| "eval_loss": 0.24872656166553497, | |
| "eval_runtime": 57.6765, | |
| "eval_samples_per_second": 16.818, | |
| "eval_steps_per_second": 2.115, | |
| "eval_wer": 0.4547212375120851, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.931506849315069, | |
| "eval_loss": 0.247171550989151, | |
| "eval_runtime": 58.1732, | |
| "eval_samples_per_second": 16.674, | |
| "eval_steps_per_second": 2.097, | |
| "eval_wer": 0.43366634439789453, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.47945205479452, | |
| "grad_norm": 0.536637544631958, | |
| "learning_rate": 0.00024754838709677417, | |
| "loss": 0.3377, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.47945205479452, | |
| "eval_loss": 0.24240365624427795, | |
| "eval_runtime": 58.1149, | |
| "eval_samples_per_second": 16.691, | |
| "eval_steps_per_second": 2.099, | |
| "eval_wer": 0.44666451820818565, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.027397260273973, | |
| "eval_loss": 0.23721589148044586, | |
| "eval_runtime": 57.9941, | |
| "eval_samples_per_second": 16.726, | |
| "eval_steps_per_second": 2.104, | |
| "eval_wer": 0.42743581480287896, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 6.575342465753424, | |
| "eval_loss": 0.23664695024490356, | |
| "eval_runtime": 58.3196, | |
| "eval_samples_per_second": 16.632, | |
| "eval_steps_per_second": 2.092, | |
| "eval_wer": 0.4224943602964873, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 6.8493150684931505, | |
| "grad_norm": 0.45388591289520264, | |
| "learning_rate": 0.00023372350230414744, | |
| "loss": 0.3282, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 7.123287671232877, | |
| "eval_loss": 0.23390649259090424, | |
| "eval_runtime": 58.129, | |
| "eval_samples_per_second": 16.687, | |
| "eval_steps_per_second": 2.099, | |
| "eval_wer": 0.4103555698786121, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 7.671232876712329, | |
| "eval_loss": 0.23517899215221405, | |
| "eval_runtime": 57.9143, | |
| "eval_samples_per_second": 16.749, | |
| "eval_steps_per_second": 2.107, | |
| "eval_wer": 0.4192716725749275, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 8.219178082191782, | |
| "grad_norm": 0.6077154278755188, | |
| "learning_rate": 0.0002198986175115207, | |
| "loss": 0.3018, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 8.219178082191782, | |
| "eval_loss": 0.2249327301979065, | |
| "eval_runtime": 57.7536, | |
| "eval_samples_per_second": 16.795, | |
| "eval_steps_per_second": 2.112, | |
| "eval_wer": 0.4097110323343001, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 8.767123287671232, | |
| "eval_loss": 0.22541025280952454, | |
| "eval_runtime": 58.012, | |
| "eval_samples_per_second": 16.721, | |
| "eval_steps_per_second": 2.103, | |
| "eval_wer": 0.40648834461274036, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 9.315068493150685, | |
| "eval_loss": 0.2250933200120926, | |
| "eval_runtime": 58.2195, | |
| "eval_samples_per_second": 16.661, | |
| "eval_steps_per_second": 2.096, | |
| "eval_wer": 0.40208400472660866, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 9.58904109589041, | |
| "grad_norm": 0.45964235067367554, | |
| "learning_rate": 0.00020607373271889397, | |
| "loss": 0.2945, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 9.863013698630137, | |
| "eval_loss": 0.22482535243034363, | |
| "eval_runtime": 57.7885, | |
| "eval_samples_per_second": 16.785, | |
| "eval_steps_per_second": 2.111, | |
| "eval_wer": 0.396927704372113, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 10.41095890410959, | |
| "eval_loss": 0.22121191024780273, | |
| "eval_runtime": 58.297, | |
| "eval_samples_per_second": 16.639, | |
| "eval_steps_per_second": 2.093, | |
| "eval_wer": 0.4001503920936728, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 10.95890410958904, | |
| "grad_norm": 1.2428817749023438, | |
| "learning_rate": 0.0001922488479262673, | |
| "loss": 0.2843, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 10.95890410958904, | |
| "eval_loss": 0.21997873485088348, | |
| "eval_runtime": 57.8696, | |
| "eval_samples_per_second": 16.762, | |
| "eval_steps_per_second": 2.108, | |
| "eval_wer": 0.39198624986572134, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 11.506849315068493, | |
| "eval_loss": 0.21829502284526825, | |
| "eval_runtime": 57.8135, | |
| "eval_samples_per_second": 16.778, | |
| "eval_steps_per_second": 2.11, | |
| "eval_wer": 0.3853260285744978, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 12.054794520547945, | |
| "eval_loss": 0.21739539504051208, | |
| "eval_runtime": 57.004, | |
| "eval_samples_per_second": 17.016, | |
| "eval_steps_per_second": 2.14, | |
| "eval_wer": 0.3889784079922656, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 12.32876712328767, | |
| "grad_norm": 0.44710573554039, | |
| "learning_rate": 0.00017842396313364056, | |
| "loss": 0.2755, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 12.602739726027398, | |
| "eval_loss": 0.21631112694740295, | |
| "eval_runtime": 57.2582, | |
| "eval_samples_per_second": 16.941, | |
| "eval_steps_per_second": 2.131, | |
| "eval_wer": 0.3955312063594371, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 13.150684931506849, | |
| "eval_loss": 0.21974815428256989, | |
| "eval_runtime": 57.1085, | |
| "eval_samples_per_second": 16.985, | |
| "eval_steps_per_second": 2.136, | |
| "eval_wer": 0.3894080996884735, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 13.698630136986301, | |
| "grad_norm": 0.3474729061126709, | |
| "learning_rate": 0.00016459907834101383, | |
| "loss": 0.2699, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 13.698630136986301, | |
| "eval_loss": 0.21634985506534576, | |
| "eval_runtime": 57.267, | |
| "eval_samples_per_second": 16.938, | |
| "eval_steps_per_second": 2.13, | |
| "eval_wer": 0.38994521430873347, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 14.246575342465754, | |
| "eval_loss": 0.21288762986660004, | |
| "eval_runtime": 56.9225, | |
| "eval_samples_per_second": 17.041, | |
| "eval_steps_per_second": 2.143, | |
| "eval_wer": 0.37694704049844235, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 14.794520547945206, | |
| "eval_loss": 0.2114371657371521, | |
| "eval_runtime": 57.0542, | |
| "eval_samples_per_second": 17.001, | |
| "eval_steps_per_second": 2.138, | |
| "eval_wer": 0.37587281125792243, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 15.068493150684931, | |
| "grad_norm": 0.5010984539985657, | |
| "learning_rate": 0.0001507741935483871, | |
| "loss": 0.2568, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 15.342465753424657, | |
| "eval_loss": 0.20998027920722961, | |
| "eval_runtime": 56.9906, | |
| "eval_samples_per_second": 17.02, | |
| "eval_steps_per_second": 2.141, | |
| "eval_wer": 0.3721130089161027, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 15.89041095890411, | |
| "eval_loss": 0.2139931619167328, | |
| "eval_runtime": 56.9487, | |
| "eval_samples_per_second": 17.033, | |
| "eval_steps_per_second": 2.142, | |
| "eval_wer": 0.36695670856160706, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 16.438356164383563, | |
| "grad_norm": 0.3882729411125183, | |
| "learning_rate": 0.00013694930875576036, | |
| "loss": 0.2521, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 16.438356164383563, | |
| "eval_loss": 0.21490569412708282, | |
| "eval_runtime": 57.0453, | |
| "eval_samples_per_second": 17.004, | |
| "eval_steps_per_second": 2.139, | |
| "eval_wer": 0.37426146739714256, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 16.986301369863014, | |
| "eval_loss": 0.21307234466075897, | |
| "eval_runtime": 57.0514, | |
| "eval_samples_per_second": 17.002, | |
| "eval_steps_per_second": 2.138, | |
| "eval_wer": 0.3720055859920507, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 16.986301369863014, | |
| "step": 6200, | |
| "total_flos": 9.649821343406506e+18, | |
| "train_loss": 0.4968759044524162, | |
| "train_runtime": 6151.3654, | |
| "train_samples_per_second": 14.226, | |
| "train_steps_per_second": 1.78 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10950, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.649821343406506e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |