Hugging Face's logo

Diffusers documentation

StableCascadeUNet

You are viewing main version, which requires installation from source. If you'd like regular pip install, checkout the latest stable version (v0.43.0).
Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

StableCascadeUNet

Un modèle UNet issu du pipeline Stable Cascade.

StableCascadeUNet

classe diffusers.models.StableCascadeUNet

< >

( in_channels: int = 16out_channels: int = 16timestep_ratio_embedding_dim: int = 64patch_size: int = 1conditioning_dim: int = 2048block_out_channels: tuple[int, ...] = (2048, 2048)num_attention_heads: tuple[int, ...] = (32, 32)down_num_layers_per_block: tuple[int, ...] = (8, 24)up_num_layers_per_block: tuple[int, ...] = (24, 8)down_blocks_repeat_mappers: tuple[int] | None = (1, 1)up_blocks_repeat_mappers: tuple[int] | None = (1, 1)block_types_per_layer: tuple[tuple[str]] = (('SDCascadeResBlock', 'SDCascadeTimestepBlock', 'SDCascadeAttnBlock'), ('SDCascadeResBlock', 'SDCascadeTimestepBlock', 'SDCascadeAttnBlock'))clip_text_in_channels: int | None = Noneclip_text_pooled_in_channels = 1280clip_image_in_channels: int | None = Noneclip_seq = 4effnet_in_channels: int | None = Nonepixel_mapper_in_channels: int | None = Nonekernel_size = 3dropout: float | tuple[float] = (0.1, 0.1)self_attn: bool | tuple[bool] = Truetimestep_conditioning_type: tuple[str, ...] = ('sca', 'crp')switch_level: tuple[bool] | None = None )

forward

< >

( sampletimestep_ratioclip_text_pooledclip_text = Noneclip_img = Noneeffnet = Nonepixels = Nonesca = Nonecrp = Nonereturn_dict = True )

Paramètres

  • sample (torch.Tensor) — L'échantillon d'entrée bruité.
  • timestep_ratio (torch.Tensor) — Ratio de pas de temps utilisé pour calculer l'embedding de pas de temps.
  • clip_text_pooled (torch.Tensor) — Embeddings de texte CLIP agrégés.
  • clip_text (torch.Tensor, facultatif) — Embeddings de texte CLIP au niveau de la séquence.
  • clip_img (torch.Tensor, facultatif) — Embeddings d’images CLIP.
  • effnet (torch.Tensor, facultatif) — Carte de caractéristiques EfficientNet utilisée comme conditionnement supplémentaire.
  • pixels (torch.Tensor, optional) — Pixel-level conditioning tensor. If None, a tensor of zeros is used.
  • sca (torch.Tensor, facultatif) — Valeur de conditionnement sca facultative utilisée pour construire l’embedding de timestep.
  • crp (torch.Tensor, facultatif) — Valeur de conditionnement crp facultative utilisée pour construire l’embedding de timestep.
  • return_dict (bool, facultatif, par défaut True) — Indique s'il faut retourner un StableCascadeUNetOutput plutôt qu'un simple tuple.
Update on GitHub

←ZImageTransformer2DModel UNet1DModel→