diffusers/pr_11739/en/api/models/autoencoder_kl_hunyuan_video15.md

AutoencoderKLHunyuanVideo15

The 3D variational autoencoder (VAE) model with KL loss used in HunyuanVideo1.5 by Tencent.

The model can be loaded with the following code snippet.

from diffusers import AutoencoderKLHunyuanVideo15

vae = AutoencoderKLHunyuanVideo15.from_pretrained("hunyuanvideo-community/HunyuanVideo-1.5-Diffusers-480p_t2v", subfolder="vae", torch_dtype=torch.float32)

# make sure to enable tiling to avoid OOM
vae.enable_tiling()

AutoencoderKLHunyuanVideo15[[diffusers.AutoencoderKLHunyuanVideo15]]

diffusers.AutoencoderKLHunyuanVideo15[[diffusers.AutoencoderKLHunyuanVideo15]]

Source

A VAE model with KL loss for encoding videos into latents and decoding latent representations into videos. Used for HunyuanVideo-1.5.

This model inherits from ModelMixin. Check the superclass documentation for it's generic methods implemented for all models (such as downloading or saving).

wrapperdiffusers.AutoencoderKLHunyuanVideo15.decodehttps://github.com/huggingface/diffusers/blob/vr_11739/src/diffusers/utils/accelerate_utils.py#L43[{"name": "*args", "val": ""}, {"name": "**kwargs", "val": ""}]

wrapper[[diffusers.AutoencoderKLHunyuanVideo15.encode]]

Source

enable_tiling[[diffusers.AutoencoderKLHunyuanVideo15.enable_tiling]]

Source

Enable tiled VAE decoding. When this option is enabled, the VAE will split the input tensor into tiles to compute decoding and encoding in several steps. This is useful for saving a large amount of memory and to allow processing larger images.

Parameters:

tile_sample_min_height (int, optional) : The minimum height required for a sample to be separated into tiles across the height dimension.

tile_sample_min_width (int, optional) : The minimum width required for a sample to be separated into tiles across the width dimension.

tile_latent_min_height (int, optional) : The minimum height required for a latent to be separated into tiles across the height dimension.

tile_latent_min_width (int, optional) : The minimum width required for a latent to be separated into tiles across the width dimension.

forward[[diffusers.AutoencoderKLHunyuanVideo15.forward]]

Source

Parameters:

sample (torch.Tensor) : Input sample.

sample_posterior (bool, optional, defaults to False) : Whether to sample from the posterior.

return_dict (bool, optional, defaults to True) : Whether or not to return a DecoderOutput instead of a plain tuple.

tiled_decode[[diffusers.AutoencoderKLHunyuanVideo15.tiled_decode]]

Source

Decode a batch of images using a tiled decoder.

Parameters:

z (torch.Tensor) : Input batch of latent vectors.

return_dict (bool, optional, defaults to True) : Whether or not to return a ~models.vae.DecoderOutput instead of a plain tuple.

Returns:

~models.vae.DecoderOutput` or `tuple

If return_dict is True, a ~models.vae.DecoderOutput is returned, otherwise a plain tuple is returned.

tiled_encode[[diffusers.AutoencoderKLHunyuanVideo15.tiled_encode]]

Source

Encode a batch of images using a tiled encoder.

Parameters:

x (torch.Tensor) : Input batch of videos.

Returns:

torch.Tensor

The latent representation of the encoded videos.

DecoderOutput[[diffusers.models.autoencoders.vae.DecoderOutput]]

diffusers.models.autoencoders.vae.DecoderOutput[[diffusers.models.autoencoders.vae.DecoderOutput]]

Source

Output of decoding method.

Parameters:

sample (torch.Tensor of shape (batch_size, num_channels, height, width)) : The decoded output sample from the last layer of the model.