diffusers

diffusers copied to clipboard

WIP: some optimizations

3

TODO: - [ ] Run an actual benchmark testing that it's faster - [ ] Maybe `Linear` doesn't make sense as running `Conv1d` would remove quite a few reshape/contiguous etc...

thomasw21

Added sampling script for unconditional training example

7

Added sampling script that supports both DDPM and DDIM sampling for unconditional training example

lukovnikov

[Community] Is there a code for pre-training stable diffusion?

1

This project works great. Is there a code for pre-training stable diffusion?

980202006

### Describe the bug When using ONNX with the custom pipeline `lpw_stable_diffusion_onnx` it raises a "unsupported operand" error. It's the same error as this other bug #967 ### Reproduction ```python...

azuritecoin

Create train_dreambooth_inpaint.py

8

train_dreambooth.py adapted to work with the inpaint model, generating random masks during the training

thedarkzeno

Music Spectrogram diffusion pipeline

14

For issue #320 and #544

kashif

[ONNX] PyTorch IO binding for faster GPU inference

2

**Context** Currently [`OnnxStableDiffusionPipeline`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/pipelines/stable_diffusion/pipeline_onnx_stable_diffusion.py#LL34C7-L34C34) perform unnecessary tensor casting between torch and numpy. The downsides of that are: * The pipeline code is harder to maintain: we have to keep track of...

anton-l

### Describe the bug I'm not able to run the Flax pipeline example of [CompVis/stable-diffusion-v1-4](https://huggingface.co/CompVis/stable-diffusion-v1-4) on Colab with GPU enabled. I get this error ``` ValueError: pmap got inconsistent sizes...

dzlab

method to prepare offline cache

7

`.from_pretrained()` can work in offline mode by loading from the cache, but we lack a method to explicitly populate that cache. I'd like something along the lines of `.cache_pretrained_for_offline()` that...

keturn

[From pretrained] Allow returning local path

1

It is a common use case to want to know exactly where the pipeline was downloaded. This PR adds a simply kwarg `return_cached_folder` to see this. It could help with...

patrickvonplaten