Causal 3d convolusion.

Causal 3d convolusion.

Simple wrapper to match the output format expected by pipeline

Causal self-attention with a single head.

Causal self-attention with a single head.

Copied from https://github.com/hehao13/CameraCtrl/blob/main/inference.py

Step Distillation Scheduler for accelerated inference. This scheduler works with step-distilled LoRA models to enable 4-step inferen

r""" Pipeline for text-to-video generation using Wan. This model inherits from [`DiffusionPipeline`]. Check the superclass documentation f

Timestep Embedding Aware Cache, a training-free caching approach that estimates and leverages the fluctuating differences among model outputs

Tile VAE Wrapper for processing high-resolution videos This implementation matches LightX2V exactly: 1. Uses low-level encoder/decod

r""" Output class for CogVideo pipelines. Args: video (`torch.Tensor`, `np.ndarray`, or List[List[PIL.Image.Image]]):

r""" Output class for CogVideo pipelines. Args: video (`torch.Tensor`, `np.ndarray`, or List[List[PIL.Image.Image]]): Lis

r""" Output class for CogVideo pipelines. Args: video (`torch.Tensor`, `np.ndarray`, or List[List[PIL.Image.Image]]): Lis

r""" Output class for CogVideo pipelines. Args: video (`torch.Tensor`, `np.ndarray`, or List[List[PIL.Image.Image]]): Lis

XLMRobertaModel with no pooler and no LM head.

`FlowDPMSolverMultistepScheduler` is a fast dedicated high-order solver for diffusion ODEs. This model inherits from [`SchedulerMixin`] and [

`UniPCMultistepScheduler` is a training-free framework designed for the fast sampling of diffusion models. This model inherits from [`Schedu

replaces forward method of the original Linear, instead of replacing the original Linear module.

Wrapper to make StepDistillScheduler compatible with existing pipeline code. This wrapper intercepts the scheduler calls and adapts them

r""" Pipeline for text-to-video generation using Wan. This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for

r""" Pipeline for text-to-video generation using Wan. This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for

r""" Pipeline for text-to-video generation using Wan. This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for

r""" Wan diffusion backbone supporting both text-to-video and image-to-video.