fastvideo.v1.pipelines.stepvideo.stepvideo_pipeline

Contents

`fastvideo.v1.pipelines.stepvideo.stepvideo_pipeline`#

Hunyuan video diffusion pipeline implementation.

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

Module Contents#

Classes#

StepVideoPipeline

Data#

`EntryClass`
`logger`

API#

fastvideo.v1.pipelines.stepvideo.stepvideo_pipeline.EntryClass#: None

class fastvideo.v1.pipelines.stepvideo.stepvideo_pipeline.StepVideoPipeline(model_path: str, fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs, config: Optional[Dict[str, Any]] = None, required_config_modules: Optional[List[str]] = None)[source]#

Bases: fastvideo.v1.pipelines.composed_pipeline_base.ComposedPipelineBase

build_clip(model_dir, device) → fastvideo.v1.models.encoders.bert.HunyuanClip[source]#

build_llm(model_dir, device) → torch.nn.Module[source]#

create_pipeline_stages(fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs)[source]#: Set up pipeline stages with proper dependency injection.

initialize_pipeline(fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs)[source]#: Initialize the pipeline.

load_modules(fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs) → Dict[str, Any][source]#: Load the modules from the config.

fastvideo.v1.pipelines.stepvideo.stepvideo_pipeline.logger[source]#: ‘init_logger(…)’