| { | |
| "best_metric": 0.20377855002880096, | |
| "best_model_checkpoint": "/scratch/skscla001/results/mms-zeroshot-300m-bembaspeech-model/checkpoint-4600", | |
| "epoch": 3.6542515811665495, | |
| "eval_steps": 200, | |
| "global_step": 5200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14054813773717498, | |
| "eval_loss": 2.2802751064300537, | |
| "eval_runtime": 82.1473, | |
| "eval_samples_per_second": 18.162, | |
| "eval_steps_per_second": 2.276, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.28109627547434995, | |
| "eval_loss": 0.2944619655609131, | |
| "eval_runtime": 82.5798, | |
| "eval_samples_per_second": 18.067, | |
| "eval_steps_per_second": 2.264, | |
| "eval_wer": 0.5100787265334371, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.35137034434293746, | |
| "grad_norm": 0.8652539849281311, | |
| "learning_rate": 0.00029720356891289033, | |
| "loss": 2.741, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.42164441321152496, | |
| "eval_loss": 0.2641391158103943, | |
| "eval_runtime": 81.2987, | |
| "eval_samples_per_second": 18.352, | |
| "eval_steps_per_second": 2.3, | |
| "eval_wer": 0.4834328229085561, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5621925509486999, | |
| "eval_loss": 0.26114964485168457, | |
| "eval_runtime": 81.2645, | |
| "eval_samples_per_second": 18.36, | |
| "eval_steps_per_second": 2.301, | |
| "eval_wer": 0.47434899212734666, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7027406886858749, | |
| "grad_norm": 1.3384147882461548, | |
| "learning_rate": 0.00029368161540267664, | |
| "loss": 0.5962, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7027406886858749, | |
| "eval_loss": 0.2608044445514679, | |
| "eval_runtime": 81.9453, | |
| "eval_samples_per_second": 18.207, | |
| "eval_steps_per_second": 2.282, | |
| "eval_wer": 0.4830867722121291, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8432888264230499, | |
| "eval_loss": 0.24766217172145844, | |
| "eval_runtime": 81.5485, | |
| "eval_samples_per_second": 18.296, | |
| "eval_steps_per_second": 2.293, | |
| "eval_wer": 0.4562678432390345, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9838369641602249, | |
| "eval_loss": 0.24071051180362701, | |
| "eval_runtime": 81.0412, | |
| "eval_samples_per_second": 18.41, | |
| "eval_steps_per_second": 2.307, | |
| "eval_wer": 0.4567004066095683, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.0541110330288124, | |
| "grad_norm": 0.5781381130218506, | |
| "learning_rate": 0.000290159661892463, | |
| "loss": 0.536, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.1243851018973998, | |
| "eval_loss": 0.23427943885326385, | |
| "eval_runtime": 82.3276, | |
| "eval_samples_per_second": 18.123, | |
| "eval_steps_per_second": 2.271, | |
| "eval_wer": 0.44121463794445887, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.264933239634575, | |
| "eval_loss": 0.23070427775382996, | |
| "eval_runtime": 82.28, | |
| "eval_samples_per_second": 18.133, | |
| "eval_steps_per_second": 2.273, | |
| "eval_wer": 0.44225279003373996, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.4054813773717498, | |
| "grad_norm": 0.5301165580749512, | |
| "learning_rate": 0.0002866377083822493, | |
| "loss": 0.5221, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.4054813773717498, | |
| "eval_loss": 0.22523300349712372, | |
| "eval_runtime": 81.337, | |
| "eval_samples_per_second": 18.343, | |
| "eval_steps_per_second": 2.299, | |
| "eval_wer": 0.4348127000605589, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.5460295151089247, | |
| "eval_loss": 0.22277304530143738, | |
| "eval_runtime": 81.5619, | |
| "eval_samples_per_second": 18.293, | |
| "eval_steps_per_second": 2.293, | |
| "eval_wer": 0.4325633705337832, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.6865776528460998, | |
| "eval_loss": 0.21623817086219788, | |
| "eval_runtime": 81.5646, | |
| "eval_samples_per_second": 18.292, | |
| "eval_steps_per_second": 2.293, | |
| "eval_wer": 0.4252963059088156, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.7568517217146873, | |
| "grad_norm": 0.5294317007064819, | |
| "learning_rate": 0.00028311575487203566, | |
| "loss": 0.5027, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.8271257905832747, | |
| "eval_loss": 0.22000491619110107, | |
| "eval_runtime": 81.31, | |
| "eval_samples_per_second": 18.35, | |
| "eval_steps_per_second": 2.3, | |
| "eval_wer": 0.4188078553508089, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.9676739283204498, | |
| "eval_loss": 0.2131248265504837, | |
| "eval_runtime": 81.3644, | |
| "eval_samples_per_second": 18.337, | |
| "eval_steps_per_second": 2.298, | |
| "eval_wer": 0.4142226836231508, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.1082220660576247, | |
| "grad_norm": 0.3984699547290802, | |
| "learning_rate": 0.000279593801361822, | |
| "loss": 0.4818, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.1082220660576247, | |
| "eval_loss": 0.22807464003562927, | |
| "eval_runtime": 81.2983, | |
| "eval_samples_per_second": 18.352, | |
| "eval_steps_per_second": 2.3, | |
| "eval_wer": 0.42806471148023184, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.2487702037947996, | |
| "eval_loss": 0.21783553063869476, | |
| "eval_runtime": 81.2747, | |
| "eval_samples_per_second": 18.357, | |
| "eval_steps_per_second": 2.301, | |
| "eval_wer": 0.41465524699368456, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.3893183415319745, | |
| "eval_loss": 0.2122805118560791, | |
| "eval_runtime": 81.1791, | |
| "eval_samples_per_second": 18.379, | |
| "eval_steps_per_second": 2.304, | |
| "eval_wer": 0.41552037373475215, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.459592410400562, | |
| "grad_norm": 0.6205651760101318, | |
| "learning_rate": 0.00027607184785160833, | |
| "loss": 0.4619, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.52986647926915, | |
| "eval_loss": 0.21417662501335144, | |
| "eval_runtime": 81.5176, | |
| "eval_samples_per_second": 18.303, | |
| "eval_steps_per_second": 2.294, | |
| "eval_wer": 0.40790725841335757, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.6704146170063248, | |
| "eval_loss": 0.21559232473373413, | |
| "eval_runtime": 81.1016, | |
| "eval_samples_per_second": 18.397, | |
| "eval_steps_per_second": 2.306, | |
| "eval_wer": 0.40107275715892376, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.8109627547434997, | |
| "grad_norm": 0.5996519923210144, | |
| "learning_rate": 0.0002725498943413947, | |
| "loss": 0.464, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.8109627547434997, | |
| "eval_loss": 0.2071918100118637, | |
| "eval_runtime": 82.6744, | |
| "eval_samples_per_second": 18.047, | |
| "eval_steps_per_second": 2.262, | |
| "eval_wer": 0.40089973181071026, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.9515108924806746, | |
| "eval_loss": 0.21284320950508118, | |
| "eval_runtime": 81.7324, | |
| "eval_samples_per_second": 18.255, | |
| "eval_steps_per_second": 2.288, | |
| "eval_wer": 0.401332295181244, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 3.0920590302178494, | |
| "eval_loss": 0.20559127628803253, | |
| "eval_runtime": 81.3555, | |
| "eval_samples_per_second": 18.339, | |
| "eval_steps_per_second": 2.299, | |
| "eval_wer": 0.3982178389134008, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 3.162333099086437, | |
| "grad_norm": 0.6319628357887268, | |
| "learning_rate": 0.000269027940831181, | |
| "loss": 0.4464, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.232607167955025, | |
| "eval_loss": 0.20377855002880096, | |
| "eval_runtime": 82.1978, | |
| "eval_samples_per_second": 18.151, | |
| "eval_steps_per_second": 2.275, | |
| "eval_wer": 0.40072670646249675, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 3.3731553056921997, | |
| "eval_loss": 0.20886844396591187, | |
| "eval_runtime": 81.607, | |
| "eval_samples_per_second": 18.283, | |
| "eval_steps_per_second": 2.291, | |
| "eval_wer": 0.39873691495804137, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 3.5137034434293746, | |
| "grad_norm": 0.6382594704627991, | |
| "learning_rate": 0.0002655059873209673, | |
| "loss": 0.4418, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.5137034434293746, | |
| "eval_loss": 0.20426709949970245, | |
| "eval_runtime": 81.6377, | |
| "eval_samples_per_second": 18.276, | |
| "eval_steps_per_second": 2.291, | |
| "eval_wer": 0.40089973181071026, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.6542515811665495, | |
| "eval_loss": 0.2049737125635147, | |
| "eval_runtime": 81.6534, | |
| "eval_samples_per_second": 18.272, | |
| "eval_steps_per_second": 2.29, | |
| "eval_wer": 0.3964010727571589, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 3.6542515811665495, | |
| "step": 5200, | |
| "total_flos": 8.128487028902814e+18, | |
| "train_loss": 0.7083046766427847, | |
| "train_runtime": 5849.6566, | |
| "train_samples_per_second": 58.347, | |
| "train_steps_per_second": 7.298 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 42690, | |
| "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": 8.128487028902814e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |