Disable the redundant torch default initialization to accelerate model creation.

r""" Instantiate a [`ControlNetModel`] from [`UNet2DConditionModel`]. Parameters: unet (`UNet2DConditionModel`):

This will load the LoRA layers specified in `state_dict` into `text_encoder` Parameters: state_dict (`dict`):

Prepares a single grid of images. Useful for visualization purposes.

Updates paths inside resnets to the new naming scheme (local renaming)

Runs `command` with `subprocess.check_output` and will potentially return the `stdout`. Will also properly capture if an error occurred while

Removes segments. Positive values shave the first segments, negative shave the last segments.

Applies the FreeU mechanism as introduced in https: //arxiv.org/abs/2309.11497. Adapted from the official code repository: https://github.com/Chen

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

Generates and yields batches from the reordered array. Parameters: - n (int): The size of each batch. Defaults to 1.

Reads a checkpoint file, returning properly formatted errors if they arise.

Convert a NumPy image to a PyTorch tensor.

r""" Removes all hooks that were added when using `enable_sequential_cpu_offload` or `enable_model_cpu_offload`.

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

This function return the height and width that are downscaled to the next integer multiple of `vae_scale_factor`. Args: