Initialize the DMD (Distribution Matching Distillation) module. This class is self-contained and compute generator and fake score los

Initialize the ODERegression module. This class is self-contained and compute generator losses in the forward pass given prec

Initialize the ODERegression module. This class is self-contained and compute generator losses in the forward pass given prec

Convert x0 prediction to flow matching's prediction. x0_pred: the x0 prediction with shape [B, C, H, W] xt: the input noisy d

we will divide the token sequence into the following format [1 latent frame] [1 latent frame] ... [1 latent frame] We use fle

we will divide the token sequence into the following format [1 latent frame] [N latent frame] ... [N latent frame] The first

Diffusion forward corruption process. Input: - clean_latent: the clean latent with shape [B*T, C, H, W] - noi

The index for the first timestep. It should be set from pipeline with `set_begin_index` method.

Convert the diffusion network's noise prediction to x0 predidction. noise: the predicted noise with shape [B, C, H, W] xt: th

Convert the diffusion network's velocity prediction to x0 predidction. velocity: the predicted noise with shape [B, C, H, W]

videos: A list of videos each with shape [C, T, H, W].

r""" Args: x(Tensor): Shape [B, L, num_heads, C / num_heads] seq_lens(Tensor): Shape [B] grid_sizes(Tensor

r""" Args: x(Tensor): Shape [B, L1, C] context(Tensor): Shape [B, L2, C] context_lens(Tensor): Shape [B]

r""" Args: x(Tensor): Shape [B, L, C] e(Tensor): Shape [B, F, 6, C] seq_lens(Tensor): Shape [B], length of

r""" Args: x(Tensor): Shape [B, L1, C] e(Tensor): Shape [B, F, 1, C]

x: [B, L1, C]. context: [B, L2, C] or None. mask: [B, L2] or [B, L1, L2] or None.

Fix bfloat16 support for nearest neighbor interpolation.

Fix bfloat16 support for nearest neighbor interpolation.

r""" Args: x(Tensor): Shape [B, L, C]

r""" Args: x(Tensor): Shape [B, L, C]

r""" Args: x(Tensor): Shape [B, L, num_heads, C / num_heads] seq_lens(Tensor): Shape [B] grid_sizes(Tensor

r""" Args: x(Tensor): Shape [B, L1, C] context(Tensor): Shape [B, L2, C] context_lens(Tensor): Shape [B]

r""" Args: x(Tensor): Shape [B, L, C] e(Tensor): Shape [B, L1, 6, C] seq_lens(Tensor): Shape [B], length o

Generate image/videos from noise and compute the DMD loss. The noisy input to the generator is backward simulated. This remov

Ensures interchangeability with schedulers that need to scale the denoising model input depending on the current timestep. A

Sets the begin index for the scheduler. This function should be run from pipeline before the inference. Args: begin_inde

The index counter for current timestep. It will increase 1 after each scheduler step.