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

r""" Converts a state dict from UNet format to diffusers format - i.e. by removing some keys

r""" Offloads all models to CPU using accelerate, reducing memory usage with a low impact on performance. Compared to `enable_sequenti

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

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

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

Upload model, scheduler, or pipeline files to the 🤗 Hugging Face Hub. Parameters: repo_id (`str`): The n

Predict the sample from the previous timestep by reversing the SDE. This function propagates the diffusion process from the learned m

Construct a layernorm module in the T5 style. No bias and no subtraction of mean.

Sets the `lora_state_dict` into `text_encoder` coming from `transformers`. Args: lora_state_dict: The state dictionary to be set.

This does the final conversion step: take locally converted weights and apply a global renaming to them. It splits attention layers, and take

Encodes the previously tokenized prompt for this diffusion region using a given encoder

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

r""" Load pretrained attention processor layers into [`UNet2DConditionModel`]. Attention processor layers have to be defined in

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

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

To run a test in a subprocess. In particular, this can avoid (GPU) memory issue. Args: test_case (`unittest.TestCase`):

r""" Save attention processor layers to a directory so that it can be reloaded with the [`~loaders.UNet2DConditionLoadersMixin.load_at

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

Gets scaled timesteps from the Karras sigmas for input to the consistency model. Args: sigmas (`float` or `np.ndarray`):

Correct the predicted sample based on the `model_output` of the network. This is often run repeatedly after making the prediction for

r""" Reverses the effect of [`pipe.fuse_lora()`](https://huggingface.co/docs/diffusers/main/en/api/loaders#diffusers.loaders.LoraLoade