fastvideo.training.wan_self_forcing_distillation_pipeline

Module Contents

Classes

WanSelfForcingDistillationPipeline

A self-forcing distillation pipeline for Wan that uses the self-forcing methodology with DMD for video generation.

Functions

main

Data

logger
vsa_available

API

class fastvideo.training.wan_self_forcing_distillation_pipeline.WanSelfForcingDistillationPipeline(model_path: str, fastvideo_args: fastvideo.fastvideo_args.TrainingArgs, required_config_modules: list[str] | None = None, loaded_modules: dict[str, torch.nn.Module] | None = None)

Bases: fastvideo.training.self_forcing_distillation_pipeline.SelfForcingDistillationPipeline

A self-forcing distillation pipeline for Wan that uses the self-forcing methodology with DMD for video generation.

Initialization

Initialize the pipeline. After init, the pipeline should be ready to use. The pipeline should be stateless and not hold any batch state.

create_training_stages(training_args: fastvideo.fastvideo_args.TrainingArgs)

May be used in future refactors.

initialize_validation_pipeline(training_args: fastvideo.fastvideo_args.TrainingArgs)

fastvideo.training.wan_self_forcing_distillation_pipeline.logger

‘init_logger(…)’

fastvideo.training.wan_self_forcing_distillation_pipeline.main(args) → None

fastvideo.training.wan_self_forcing_distillation_pipeline.vsa_available

‘is_vsa_available(…)’