| { | |
| "best_metric": 0.2293323278427124, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-swagen-combined-15hrs-model/checkpoint-900", | |
| "epoch": 1.035870864886409, | |
| "eval_steps": 100, | |
| "global_step": 1300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07971303308090873, | |
| "grad_norm": 4.197810173034668, | |
| "learning_rate": 0.000279, | |
| "loss": 14.8801, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07971303308090873, | |
| "eval_loss": 0.7376943230628967, | |
| "eval_runtime": 64.6778, | |
| "eval_samples_per_second": 17.502, | |
| "eval_steps_per_second": 4.376, | |
| "eval_wer": 0.44259038537941997, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15942606616181745, | |
| "grad_norm": 4.988656997680664, | |
| "learning_rate": 0.0002992563965884861, | |
| "loss": 0.6766, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.15942606616181745, | |
| "eval_loss": 0.2687693238258362, | |
| "eval_runtime": 64.5932, | |
| "eval_samples_per_second": 17.525, | |
| "eval_steps_per_second": 4.381, | |
| "eval_wer": 0.20063567739372268, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2391390992427262, | |
| "grad_norm": 3.365363836288452, | |
| "learning_rate": 0.0002984568230277185, | |
| "loss": 0.5153, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2391390992427262, | |
| "eval_loss": 0.24840499460697174, | |
| "eval_runtime": 64.6789, | |
| "eval_samples_per_second": 17.502, | |
| "eval_steps_per_second": 4.375, | |
| "eval_wer": 0.19745729042510926, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3188521323236349, | |
| "grad_norm": 2.4677116870880127, | |
| "learning_rate": 0.00029765724946695095, | |
| "loss": 0.526, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3188521323236349, | |
| "eval_loss": 0.23976168036460876, | |
| "eval_runtime": 64.8543, | |
| "eval_samples_per_second": 17.455, | |
| "eval_steps_per_second": 4.364, | |
| "eval_wer": 0.19487485101311083, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3985651654045436, | |
| "grad_norm": 3.4616479873657227, | |
| "learning_rate": 0.0002968576759061834, | |
| "loss": 0.4874, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3985651654045436, | |
| "eval_loss": 0.23978756368160248, | |
| "eval_runtime": 65.3676, | |
| "eval_samples_per_second": 17.317, | |
| "eval_steps_per_second": 4.329, | |
| "eval_wer": 0.19576877234803336, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4782781984854524, | |
| "grad_norm": 1.9864723682403564, | |
| "learning_rate": 0.00029605810234541576, | |
| "loss": 0.4666, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4782781984854524, | |
| "eval_loss": 0.2357860654592514, | |
| "eval_runtime": 64.7957, | |
| "eval_samples_per_second": 17.47, | |
| "eval_steps_per_second": 4.368, | |
| "eval_wer": 0.19090186730234407, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5579912315663611, | |
| "grad_norm": 2.5512609481811523, | |
| "learning_rate": 0.00029525852878464813, | |
| "loss": 0.4406, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5579912315663611, | |
| "eval_loss": 0.2390868365764618, | |
| "eval_runtime": 64.7802, | |
| "eval_samples_per_second": 17.474, | |
| "eval_steps_per_second": 4.369, | |
| "eval_wer": 0.194378228049265, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6377042646472698, | |
| "grad_norm": 6.3221330642700195, | |
| "learning_rate": 0.00029445895522388056, | |
| "loss": 0.4689, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6377042646472698, | |
| "eval_loss": 0.23335325717926025, | |
| "eval_runtime": 64.7754, | |
| "eval_samples_per_second": 17.476, | |
| "eval_steps_per_second": 4.369, | |
| "eval_wer": 0.19259038537941994, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7174172977281785, | |
| "grad_norm": 2.1293585300445557, | |
| "learning_rate": 0.000293659381663113, | |
| "loss": 0.462, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7174172977281785, | |
| "eval_loss": 0.2293323278427124, | |
| "eval_runtime": 65.6055, | |
| "eval_samples_per_second": 17.255, | |
| "eval_steps_per_second": 4.314, | |
| "eval_wer": 0.19268970997218912, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7971303308090872, | |
| "grad_norm": 6.672135353088379, | |
| "learning_rate": 0.00029285980810234537, | |
| "loss": 0.4407, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7971303308090872, | |
| "eval_loss": 0.22934316098690033, | |
| "eval_runtime": 65.0636, | |
| "eval_samples_per_second": 17.398, | |
| "eval_steps_per_second": 4.35, | |
| "eval_wer": 0.19308700834326578, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8768433638899961, | |
| "grad_norm": 7.573569297790527, | |
| "learning_rate": 0.0002920602345415778, | |
| "loss": 0.4567, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8768433638899961, | |
| "eval_loss": 0.22979336977005005, | |
| "eval_runtime": 65.234, | |
| "eval_samples_per_second": 17.353, | |
| "eval_steps_per_second": 4.338, | |
| "eval_wer": 0.19278903456495827, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9565563969709048, | |
| "grad_norm": 2.2756216526031494, | |
| "learning_rate": 0.00029126066098081023, | |
| "loss": 0.4711, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9565563969709048, | |
| "eval_loss": 0.23050223290920258, | |
| "eval_runtime": 65.3786, | |
| "eval_samples_per_second": 17.315, | |
| "eval_steps_per_second": 4.329, | |
| "eval_wer": 0.19715931664680175, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.035870864886409, | |
| "grad_norm": 1.712958574295044, | |
| "learning_rate": 0.0002904610874200426, | |
| "loss": 0.4444, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.035870864886409, | |
| "eval_loss": 0.23069703578948975, | |
| "eval_runtime": 66.0724, | |
| "eval_samples_per_second": 17.133, | |
| "eval_steps_per_second": 4.283, | |
| "eval_wer": 0.19288835915772745, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.035870864886409, | |
| "step": 1300, | |
| "total_flos": 5.341353862310001e+18, | |
| "train_loss": 1.5951181411743165, | |
| "train_runtime": 2475.2222, | |
| "train_samples_per_second": 121.638, | |
| "train_steps_per_second": 15.199 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 37620, | |
| "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": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.341353862310001e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |