biddlmZddlmZmZmZddlZddlmZddl m Z m Z ddl m Z ddlmZdd lmZmZmZmZdd lmZeGd d e ZGd dee Zy)) dataclass)OptionalTupleUnionN) ConfigMixinregister_to_config) BaseOutput)apply_forward_hook)Decoder DecoderOutputEncoderVectorQuantizer) ModelMixinc0eZdZUdZej ed<y)VQEncoderOutputz Output of VQModel encoding method. Args: latents (`torch.Tensor` of shape `(batch_size, num_channels, height, width)`): The encoded output sample from the last layer of the model. latentsN)__name__ __module__ __qualname____doc__torchTensor__annotations__a/home/cdr/jupyterlab/.venv/lib/python3.12/site-packages/diffusers/models/autoencoders/vq_model.pyrrs\\rrceZdZdZdgZdZe ddededee dfdee dfd eedfd ed e d ed ededede ede de ffd Z e ddejdedefdZe ddejdededeeejffdZ ddejdedeeeejdfffdZxZS) VQModelaJ A VQ-VAE model for decoding latent representations. This model inherits from [`ModelMixin`]. Check the superclass documentation for it's generic methods implemented for all models (such as downloading or saving). Parameters: in_channels (int, *optional*, defaults to 3): Number of channels in the input image. out_channels (int, *optional*, defaults to 3): Number of channels in the output. down_block_types (`Tuple[str]`, *optional*, defaults to `("DownEncoderBlock2D",)`): Tuple of downsample block types. up_block_types (`Tuple[str]`, *optional*, defaults to `("UpDecoderBlock2D",)`): Tuple of upsample block types. block_out_channels (`Tuple[int]`, *optional*, defaults to `(64,)`): Tuple of block output channels. layers_per_block (`int`, *optional*, defaults to `1`): Number of layers per block. act_fn (`str`, *optional*, defaults to `"silu"`): The activation function to use. latent_channels (`int`, *optional*, defaults to `3`): Number of channels in the latent space. sample_size (`int`, *optional*, defaults to `32`): Sample input size. num_vq_embeddings (`int`, *optional*, defaults to `256`): Number of codebook vectors in the VQ-VAE. norm_num_groups (`int`, *optional*, defaults to `32`): Number of groups for normalization layers. vq_embed_dim (`int`, *optional*): Hidden dim of codebook vectors in the VQ-VAE. scaling_factor (`float`, *optional*, defaults to `0.18215`): The component-wise standard deviation of the trained latent space computed using the first batch of the training set. This is used to scale the latent space to have unit variance when training the diffusion model. The latents are scaled with the formula `z = z * scaling_factor` before being passed to the diffusion model. When decoding, the latents are scaled back to the original scale with the formula: `z = 1 / scaling_factor * z`. For more details, refer to sections 4.3.2 and D.1 of the [High-Resolution Image Synthesis with Latent Diffusion Models](https://huggingface.co/papers/2112.10752) paper. norm_type (`str`, *optional*, defaults to `"group"`): Type of normalization layer to use. Can be one of `"group"` or `"spatial"`. quantizeF in_channels out_channelsdown_block_types.up_block_typesblock_out_channelslayers_per_blockact_fnlatent_channels sample_sizenum_vq_embeddingsnorm_num_groups vq_embed_dimscaling_factor norm_typec ,t|t||||||| d| |_| | n|} t j || d|_t| | ddd|_t j | |d|_ t||||||| || |_ y)NF) r"r#r$r&r'r(r,double_zmid_block_add_attentiong?)betaremapsane_index_shape) r"r#r%r&r'r(r,r/r2) super__init__rencodernnConv2d quant_convrr!post_quant_convr decoder)selfr"r#r$r%r&r'r(r)r*r+r,r-r.r/r2lookup_from_codebook force_upcast __class__s rr8zVQModel.__init__Ms* #(-1-+$;   (4'?|_ ))O\1E'(9r/r) r?rGrIrDrRquantrN_quant2decs rdecodezVQModel.decodes "$(MM!$4 !E; [[ - -MM44Q>E++qwwqz366qxxqww6OKE++qwwqz366qxxqww6OK%%e,ll6DKK,A,AY,N5TXY # #C[AArrMc|j|j}|j|}|s|j|jfS|S)a' The [`VQModel`] forward method. Args: sample (`torch.Tensor`): Input sample. return_dict (`bool`, *optional*, defaults to `True`): Whether or not to return a [`models.autoencoders.vq_model.VQEncoderOutput`] instead of a plain tuple. Returns: [`~models.autoencoders.vq_model.VQEncoderOutput`] or `tuple`: If return_dict is True, a [`~models.autoencoders.vq_model.VQEncoderOutput`] is returned, otherwise a plain `tuple` is returned. )rHrrYrMrN)r?rMrDrGrXs rforwardzVQModel.forwards@" KK  ' 'kk!n::s. . r)rr)DownEncoderBlock2D)UpDecoderBlock2D)@r3silur r`Ng{P?groupTFF)T)FTN)rrrr _skip_layerwise_casting_patterns_supports_group_offloadingr intrstrrfloatr8r rrboolrrHrrrYr[ __classcell__)rBs@rr r (sB)3|$!&,C*?.3 ! !$!&* ' $"%4 4 4  S/ 4 c3h 4 "#s(O 4 4 4 4 4 4 4 sm4 4 4 4 l* *4*?**aeBB37BNRB }ell* +BB*9=ll15 }eELL#$566 7rr ) dataclassesrtypingrrrrtorch.nnr:configuration_utilsrr utilsr utils.accelerate_utilsr autoencoders.vaer rrrmodeling_utilsrrr rrrrrsS")) B8OO'  j   Qj+Qr