A wrapper around :function:`torch.nn.functional.grid_sample` to support 3D point_coords tensors. Unlike :function:`torch.nn.functional.grid_s

r""" Combines a latent iamge img_vae of shape (B, C, H, W) and a CLIP-embedded image img_clip of shape (B, 1, clip_img_dim) into a sin

r""" Combines a latent image img_vae of shape (B, C, H, W), a CLIP-embedded image img_clip of shape (B, L_img, clip_img_dim), and a te

r""" Splits a flattened embedding x of shape (B, C * H * W + clip_img_dim + clip_img_dim] into (img_vae, img_clip, label_clip) where i

r""" Splits a flattened embedding x of shape (B, C * H * W + clip_img_dim) into two tensors of shape (B, C, H, W) and (B, 1, clip_img_

Sample points in [0, 1] x [0, 1] coordinate space based on their uncertainty. The unceratinties are calculated for each point using 'unce

Function to convert targets to the format expected by the loss Args: targets: list[Dict] ignore_label: int

r""" Gets the noise prediction using the `unet` and performs classifier-free guidance, if necessary.

Calculates the uncertainty when using sigmoid loss. Defined according to PointRend: https://arxiv.org/abs/1912.08193 Args:

Calculates the uncertainty when using a CE loss. Defined according to PointRend: https://arxiv.org/abs/1912.08193

Compute the DICE loss, similar to generalized IOU for masks Args: inputs: A float tensor of arbitrary shape.

r""" Encodes the prompt into text encoder hidden states. Args: prompt (`str` or `List[str]`, *optional*):

Uncertainty loss for instance segmentation as used in Mask2Former: https://arxiv.org/abs/2112.01527 Only minor modifications (i.e., s

Matcher comes from Mask2Former: https://arxiv.org/abs/2112.01527 This function is not used by default.

Compute the DICE loss, similar to generalized IOU for masks Args: inputs: A float tensor of arbitrary shape.

Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A flo

We use very effective losses to quantify the quality of the reconstruction. Overall loss function consists of 3 terms: -

Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A flo

r""" The call function to the pipeline for generation. Args: height (`int`, *optional*, defaults to `self.unet.config.sam

r""" Disable sliced VAE decoding. If `enable_vae_slicing` was previously enabled, this method will go back to computing decoding in on

r""" Disable tiled VAE decoding. If `enable_vae_tiling` was previously enabled, this method will go back to computing decoding in one

r""" Enable sliced VAE decoding. When this option is enabled, the VAE will split the input tensor in slices to compute decoding in sev

r""" Enable tiled VAE decoding. When this option is enabled, the VAE will split the input tensor into tiles to compute decoding and en

Args: prompt_embeds (`torch.Tensor` of shape `(batch size, seq_len, text_dim)`): CLIP-embedded text representatio