diffusers/pr_14401/en/api/models/autoencoderkl_audio_ltx_2.md

AutoencoderKLLTX2Audio

The 3D variational autoencoder (VAE) model with KL loss used in LTX-2 was introduced by Lightricks. This is for encoding and decoding audio latent representations.

The model can be loaded with the following code snippet.

from diffusers import AutoencoderKLLTX2Audio

vae = AutoencoderKLLTX2Audio.from_pretrained("Lightricks/LTX-2", subfolder="vae", dtype=torch.float32).to("cuda")

AutoencoderKLLTX2Audio[[diffusers.AutoencoderKLLTX2Audio]]

diffusers.AutoencoderKLLTX2Audio[[diffusers.AutoencoderKLLTX2Audio]]

diffusers.AutoencoderKLLTX2Audio(base_channels: int = 128, output_channels: int = 2, ch_mult: tuple = (1, 2, 4), num_res_blocks: int = 2, attn_resolutions: tuple[int, ...] | None = None, in_channels: int = 2, resolution: int = 256, latent_channels: int = 8, norm_type: str = 'pixel', causality_axis: str | None = 'height', dropout: float = 0.0, mid_block_add_attention: bool = False, sample_rate: int = 16000, mel_hop_length: int = 160, is_causal: bool = True, mel_bins: int | None = 64, double_z: bool = True)

Source

LTX2 audio VAE for encoding and decoding audio latent representations.

encode[[diffusers.AutoencoderKLLTX2Audio.encode]]

encode(x: Tensor, return_dict: bool = True)

Source

decode[[diffusers.AutoencoderKLLTX2Audio.decode]]

decode(z: Tensor, return_dict: bool = True)

Source

forward[[diffusers.AutoencoderKLLTX2Audio.forward]]

forward(sample: Tensor, sample_posterior: bool = False, return_dict: bool = True, generator: typing.Optional[torch.Generator] = None)

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.

generator (torch.Generator, optional) : A torch.Generator to make sampling deterministic.

Returns: DecoderOutput or tuple

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