| { | |
| "best_metric": 0.30884110927581787, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-lozgen-male-model/checkpoint-1800", | |
| "epoch": 16.03053435114504, | |
| "eval_steps": 100, | |
| "global_step": 2100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.7633587786259542, | |
| "grad_norm": 2.278857707977295, | |
| "learning_rate": 0.000285, | |
| "loss": 6.7665, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7633587786259542, | |
| "eval_loss": 3.1609649658203125, | |
| "eval_runtime": 20.525, | |
| "eval_samples_per_second": 14.811, | |
| "eval_steps_per_second": 3.703, | |
| "eval_wer": 0.9990169574834111, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5267175572519083, | |
| "grad_norm": 2.781928062438965, | |
| "learning_rate": 0.00029255874673629243, | |
| "loss": 2.6476, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.5267175572519083, | |
| "eval_loss": 2.2410073280334473, | |
| "eval_runtime": 20.3251, | |
| "eval_samples_per_second": 14.957, | |
| "eval_steps_per_second": 3.739, | |
| "eval_wer": 0.8953059719832883, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.2900763358778624, | |
| "grad_norm": 2.653702974319458, | |
| "learning_rate": 0.00028472584856396865, | |
| "loss": 1.5694, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.2900763358778624, | |
| "eval_loss": 0.5584537386894226, | |
| "eval_runtime": 20.2001, | |
| "eval_samples_per_second": 15.049, | |
| "eval_steps_per_second": 3.762, | |
| "eval_wer": 0.7598918653231752, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.053435114503817, | |
| "grad_norm": 1.1257972717285156, | |
| "learning_rate": 0.0002768929503916449, | |
| "loss": 0.6761, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.053435114503817, | |
| "eval_loss": 0.4565717875957489, | |
| "eval_runtime": 20.2697, | |
| "eval_samples_per_second": 14.998, | |
| "eval_steps_per_second": 3.749, | |
| "eval_wer": 0.6318505775374785, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.816793893129771, | |
| "grad_norm": 0.982060968875885, | |
| "learning_rate": 0.0002690600522193211, | |
| "loss": 0.5793, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.816793893129771, | |
| "eval_loss": 0.4013397693634033, | |
| "eval_runtime": 20.4119, | |
| "eval_samples_per_second": 14.893, | |
| "eval_steps_per_second": 3.723, | |
| "eval_wer": 0.4821823543868272, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.580152671755725, | |
| "grad_norm": 0.9821244478225708, | |
| "learning_rate": 0.0002612271540469974, | |
| "loss": 0.5392, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.580152671755725, | |
| "eval_loss": 0.3795173168182373, | |
| "eval_runtime": 20.3646, | |
| "eval_samples_per_second": 14.928, | |
| "eval_steps_per_second": 3.732, | |
| "eval_wer": 0.4553944458097813, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.34351145038168, | |
| "grad_norm": 1.2693768739700317, | |
| "learning_rate": 0.0002533942558746736, | |
| "loss": 0.4809, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.34351145038168, | |
| "eval_loss": 0.37301722168922424, | |
| "eval_runtime": 20.1691, | |
| "eval_samples_per_second": 15.073, | |
| "eval_steps_per_second": 3.768, | |
| "eval_wer": 0.4332759891865323, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.106870229007634, | |
| "grad_norm": 1.0504028797149658, | |
| "learning_rate": 0.0002455613577023499, | |
| "loss": 0.4813, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.106870229007634, | |
| "eval_loss": 0.35973772406578064, | |
| "eval_runtime": 20.1824, | |
| "eval_samples_per_second": 15.063, | |
| "eval_steps_per_second": 3.766, | |
| "eval_wer": 0.4229540427623495, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.870229007633588, | |
| "grad_norm": 1.0751953125, | |
| "learning_rate": 0.0002377284595300261, | |
| "loss": 0.4484, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.870229007633588, | |
| "eval_loss": 0.3431786596775055, | |
| "eval_runtime": 20.3009, | |
| "eval_samples_per_second": 14.975, | |
| "eval_steps_per_second": 3.744, | |
| "eval_wer": 0.39247972474809534, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.633587786259542, | |
| "grad_norm": 3.291668176651001, | |
| "learning_rate": 0.00022989556135770233, | |
| "loss": 0.4418, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.633587786259542, | |
| "eval_loss": 0.3390863239765167, | |
| "eval_runtime": 20.3166, | |
| "eval_samples_per_second": 14.963, | |
| "eval_steps_per_second": 3.741, | |
| "eval_wer": 0.39469157041042024, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.396946564885496, | |
| "grad_norm": 1.688865303993225, | |
| "learning_rate": 0.00022206266318537858, | |
| "loss": 0.4322, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 8.396946564885496, | |
| "eval_loss": 0.33392783999443054, | |
| "eval_runtime": 20.3067, | |
| "eval_samples_per_second": 14.97, | |
| "eval_steps_per_second": 3.743, | |
| "eval_wer": 0.3841238633570902, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.16030534351145, | |
| "grad_norm": 1.2780994176864624, | |
| "learning_rate": 0.0002142297650130548, | |
| "loss": 0.3963, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 9.16030534351145, | |
| "eval_loss": 0.3294471502304077, | |
| "eval_runtime": 20.282, | |
| "eval_samples_per_second": 14.989, | |
| "eval_steps_per_second": 3.747, | |
| "eval_wer": 0.36692061931678543, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 9.923664122137405, | |
| "grad_norm": 1.1817948818206787, | |
| "learning_rate": 0.00020639686684073108, | |
| "loss": 0.4104, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 9.923664122137405, | |
| "eval_loss": 0.321709543466568, | |
| "eval_runtime": 20.3324, | |
| "eval_samples_per_second": 14.952, | |
| "eval_steps_per_second": 3.738, | |
| "eval_wer": 0.3634799705087245, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 10.68702290076336, | |
| "grad_norm": 2.474210262298584, | |
| "learning_rate": 0.0001985639686684073, | |
| "loss": 0.3777, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 10.68702290076336, | |
| "eval_loss": 0.3176502287387848, | |
| "eval_runtime": 20.3222, | |
| "eval_samples_per_second": 14.959, | |
| "eval_steps_per_second": 3.74, | |
| "eval_wer": 0.3610223642172524, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 11.450381679389313, | |
| "grad_norm": 1.456008791923523, | |
| "learning_rate": 0.00019073107049608353, | |
| "loss": 0.3785, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 11.450381679389313, | |
| "eval_loss": 0.32359373569488525, | |
| "eval_runtime": 20.3296, | |
| "eval_samples_per_second": 14.954, | |
| "eval_steps_per_second": 3.738, | |
| "eval_wer": 0.35389530597198327, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 12.213740458015268, | |
| "grad_norm": 1.4315876960754395, | |
| "learning_rate": 0.00018289817232375978, | |
| "loss": 0.3682, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 12.213740458015268, | |
| "eval_loss": 0.31440281867980957, | |
| "eval_runtime": 20.1701, | |
| "eval_samples_per_second": 15.072, | |
| "eval_steps_per_second": 3.768, | |
| "eval_wer": 0.34676824772671416, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 12.977099236641221, | |
| "grad_norm": 2.2636640071868896, | |
| "learning_rate": 0.000175065274151436, | |
| "loss": 0.3654, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 12.977099236641221, | |
| "eval_loss": 0.31223368644714355, | |
| "eval_runtime": 20.3687, | |
| "eval_samples_per_second": 14.925, | |
| "eval_steps_per_second": 3.731, | |
| "eval_wer": 0.35291226345539445, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 13.740458015267176, | |
| "grad_norm": 0.9211711883544922, | |
| "learning_rate": 0.00016723237597911225, | |
| "loss": 0.3509, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 13.740458015267176, | |
| "eval_loss": 0.30884110927581787, | |
| "eval_runtime": 20.3386, | |
| "eval_samples_per_second": 14.947, | |
| "eval_steps_per_second": 3.737, | |
| "eval_wer": 0.3462767264684198, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 14.50381679389313, | |
| "grad_norm": 2.6637320518493652, | |
| "learning_rate": 0.0001593994778067885, | |
| "loss": 0.3412, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 14.50381679389313, | |
| "eval_loss": 0.314583420753479, | |
| "eval_runtime": 20.1788, | |
| "eval_samples_per_second": 15.065, | |
| "eval_steps_per_second": 3.766, | |
| "eval_wer": 0.33472597689850087, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 15.267175572519085, | |
| "grad_norm": 0.9389367699623108, | |
| "learning_rate": 0.00015156657963446475, | |
| "loss": 0.3344, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 15.267175572519085, | |
| "eval_loss": 0.31076034903526306, | |
| "eval_runtime": 20.2164, | |
| "eval_samples_per_second": 15.037, | |
| "eval_steps_per_second": 3.759, | |
| "eval_wer": 0.34160727451462275, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 16.03053435114504, | |
| "grad_norm": 0.9784950017929077, | |
| "learning_rate": 0.00014373368146214098, | |
| "loss": 0.3351, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 16.03053435114504, | |
| "eval_loss": 0.3107351064682007, | |
| "eval_runtime": 20.3863, | |
| "eval_samples_per_second": 14.912, | |
| "eval_steps_per_second": 3.728, | |
| "eval_wer": 0.3263701155074957, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 16.03053435114504, | |
| "step": 2100, | |
| "total_flos": 5.844277956095999e+18, | |
| "train_loss": 0.8914755775814964, | |
| "train_runtime": 1802.4751, | |
| "train_samples_per_second": 8.721, | |
| "train_steps_per_second": 2.18 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 3930, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "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": 5.844277956095999e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |