fastvideo.v1.pipelines.wan.wan_pipeline

Wan video diffusion pipeline implementation.

This module contains an implementation of the Wan video diffusion pipeline using the modular pipeline architecture.

Module Contents

Classes

WanPipeline	Wan video diffusion pipeline with LoRA support.
WanValidationPipeline	Validation pipeline for Wan2.1, assumes that the input are preprocess latents.

Data

EntryClass
logger

API

fastvideo.v1.pipelines.wan.wan_pipeline.EntryClass

None

class fastvideo.v1.pipelines.wan.wan_pipeline.WanPipeline(*args, **kwargs)

Bases: fastvideo.v1.pipelines.LoRAPipeline, fastvideo.v1.pipelines.ComposedPipelineBase

Wan video diffusion pipeline with LoRA support.

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_pipeline_stages(fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs) → None

Set up pipeline stages with proper dependency injection.

initialize_pipeline(fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs)

class fastvideo.v1.pipelines.wan.wan_pipeline.WanValidationPipeline(model_path: str, fastvideo_args: Union[fastvideo.v1.fastvideo_args.FastVideoArgs, fastvideo.v1.fastvideo_args.TrainingArgs], required_config_modules: Optional[List[str]] = None, loaded_modules: Optional[Dict[str, torch.nn.Module]] = None)

Bases: fastvideo.v1.pipelines.ComposedPipelineBase

Validation pipeline for Wan2.1, assumes that the input are preprocess latents.

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_pipeline_stages(fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs) → None

Set up pipeline stages with proper dependency injection.

initialize_pipeline(fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs)

fastvideo.v1.pipelines.wan.wan_pipeline.logger

‘init_logger(…)’