r"""Pad, upsample, filter, and downsample a batch of 2D images. Performs the following sequence of operations for each channel: 1. Upsample

Wrapper for the underlying `conv2d()` and `conv_transpose2d()` implementations.

r"""Convenience function to setup 2D FIR filter for `upfirdn2d()`. Args: f: Torch tensor, numpy array, or python list of the sh

r"""Fused bias and activation function. Adds bias `b` to activation tensor `x`, evaluates activation function `act`, and scales the result by

r"""2D convolution with optional up/downsampling. Padding is performed only once at the beginning, not between the operations. Args:

Args: x: (B, H, W, C) window_size (int): window size Returns: windows: (num_windows*B, window_size, window_size, C)

Args: image: [H, W, C] RGB mask: [H, W, 1] box: [left,top,right,bottom] Returns: BG

Args: mask: (h, w, 1) 0~255 Returns:

Args: img: [H, W, C] mod: square: 是否为正方形 min_size: Returns:

r"""Upsample a batch of 2D images using the given 2D FIR filter. By default, the result is padded so that its shape is a multiple of the input.

Args: windows: (num_windows*B, window_size, window_size, C) window_size (int): Window size H (int): Height of image

Slow reference implementation of `bias_act()` using standard TensorFlow ops.

Args: image: [H, W, C] RGB mask: [H, W, 1] box: [left,top,right,bottom] Returns: BG

Slow reference implementation of `upfirdn2d()` using standard PyTorch ops.

Slow reference implementation of `upfirdn2d()` using standard PyTorch ops.

r"""Downsample a batch of 2D images using the given 2D FIR filter. By default, the result is padded so that its shape is a fraction of the input.

r""" Enable sliced attention computation. When this option is enabled, the attention module will split the input tensor in slices, to

Input images and output images have same size images: [H, W, C] RGB masks: [H, W, 1] 255 为 masks 区域 return: BGR IMAGE

Args: img: [H, W, C] RGB mask: [H, W] 255 为 masks 区域 sigma256: Returns:

Create sinusoidal timestep embeddings. :param timesteps: a 1-D Tensor of N indices, one per batch element. These may be

r""" Function invoked when calling the pipeline for generation. Args: prompt (`str` or `List[str]`): The

images: [H, W, C] RGB, not normalized masks: [H, W] return: BGR IMAGE

images: [H, W, C] RGB, not normalized masks: [H, W] return: BGR IMAGE

images: [H, W, C] RGB, not normalized masks: [H, W] return: BGR IMAGE