espnet2.train.uasr_trainer.UASRTrainerOptions

source

class espnet2.train.uasr_trainer.UASRTrainerOptions(ngpu: int, resume: bool, use_amp: bool, train_dtype: str, grad_noise: bool, accum_grad: int, grad_clip: float, grad_clip_type: float, log_interval: int | None, no_forward_run: bool, use_matplotlib: bool, use_tensorboard: bool, use_wandb: bool, adapter: str, use_adapter: bool, save_strategy: str, output_dir: Path | str, max_epoch: int, seed: int, sharded_ddp: bool, patience: int | None, keep_nbest_models: int | List[int], nbest_averaging_interval: int, early_stopping_criterion: Sequence[str], best_model_criterion: Sequence[Sequence[str]], val_scheduler_criterion: Sequence[str], unused_parameters: bool, wandb_model_log_interval: int, create_graph_in_tensorboard: bool, generator_first: bool, max_num_warning: int)

Bases: TrainerOptions

Trainer option dataclass for UASRTrainer.

This class defines the options that can be configured for the UASRTrainer. It extends the TrainerOptions class and includes specific attributes relevant to the UASR training process.

generator_first

A flag indicating whether to update the generator first during training.

• Type: bool

max_num_warning

The maximum number of warnings that can be shown during training.

• Type: int

• Parameters:

  • generator_first (bool) – A flag indicating whether to update the generator first during training. Defaults to False.
  • max_num_warning (int) – The maximum number of warnings that can be shown during training. Defaults to 10.

Examples

To create an instance of UASRTrainerOptions:

python options = UASRTrainerOptions(generator_first=True, max_num_warning=5)

NOTE

The UASRTrainerOptions class is specifically designed to be used with the UASRTrainer and its training process.

generator_first

*: bool*

max_num_warning

*: int*