diffusers

diffusers copied to clipboard

### Describe the bug fp16 does not work with euler_a scheduler: it displays nearly black image. using checkpoint: stable-diffusion-v1.5 ### Reproduction - Input prompt: a boy. - euler_a scheduler: default...

KingStorm

`deprecate` should point to the usage of the deprecated thing

1

This warning message: > /opt/hostedtoolcache/Python/3.10.8/x64/lib/python3.10/site-packages/diffusers/utils/deprecation_utils.py:35: FutureWarning: It is deprecated to pass a pretrained model name or path to `from_config`.If you were trying to load a scheduler, please use .from_pretrained(...) instead....

keturn

EulerDiscreteScheduler is not complete and sign of c_out is wrong

1

### Describe the bug Hi, the implementation of EDM method here https://github.com/huggingface/diffusers/blob/main/src/diffusers/schedulers/scheduling_euler_discrete.py misses the second-order correction step implemented here in the original paper https://arxiv.org/pdf/2206.00364.pdf: please see https://github.com/NVlabs/edm/blob/b2a26c921c5776cb52f7498248761d60649007a8/example.py#L65 ``` # Apply...

cameronalonso2

Annealed Langevin Sampler

8

Closes #84 Maybe @MinkaiXu can give a review on this?

natolambert

Clean up 1d unet blocks

9

Some changes motivated by #105.

natolambert

Add INT8 Stable Diffusion through Optimum

6

8-bit quantization is useful to improve the inference performance. This PR is to add INT8 quantization for Stable Diffusion through Optimum-Intel quantization API on top of Intel Neural Compressor. The...

hshen14

### Describe the bug Using the `StableDiffusionUpscalePipeline` when called to output more than one image will crash. This occurs when using the `num_images_per_prompt` parameter or providing lists with multiple elements...

zookae

### Describe the bug The error was thrown when executing the following codes on SD V2 inpainting . It works fine on SD v1. My cuda and torch version are...

alfredcs

Fix training docs to install datasets

1

#1463 Just added documentation to add `training` packages.

pedrogengo

Add from_pretrained telemetry

3

This telemetry will mostly just allow us to gauge the proportion of local vs hub pipeline loading. No potentially sensitive info about the (custom) pipeline class or the model path...

anton-l