fastvideo.v1.pipelines.stages.encoding
#
Encoding stage for diffusion pipelines.
Module Contents#
Classes#
Stage for encoding pixel representations into latent space. |
Data#
API#
- class fastvideo.v1.pipelines.stages.encoding.EncodingStage(vae: fastvideo.v1.models.vaes.common.ParallelTiledVAE)[source]#
Bases:
fastvideo.v1.pipelines.stages.base.PipelineStage
Stage for encoding pixel representations into latent space.
This stage handles the encoding of pixel representations into the final input format (e.g., latents).
Initialization
- forward(batch: fastvideo.v1.pipelines.pipeline_batch_info.ForwardBatch, fastvideo_args: fastvideo.v1.fastvideo_args.FastVideoArgs) fastvideo.v1.pipelines.pipeline_batch_info.ForwardBatch [source]#
Encode pixel representations into latent space.
- Parameters:
batch β The current batch information.
fastvideo_args β The inference arguments.
- Returns:
The batch with encoded outputs.
- preprocess(image: PIL.Image.Image, vae_scale_factor: int, height: Optional[int] = None, width: Optional[int] = None, resize_mode: str = 'default') torch.Tensor [source]#
- retrieve_latents(encoder_output: torch.Tensor, generator: Optional[torch.Generator] = None, sample_mode: str = 'sample')[source]#