diffusers

diffusers copied to clipboard

Flax support for Stable Diffusion 2

5

pcuenca

[ONNX] Support for nightly versions of the directml package

9

Self explanatory title. The latest nightly version has improved my inference speeds considerably and I know others who use it as well but currently it fails to import because import_utils.py...

GreenLandisaLie

[wip] kakaobrain unCLIP

4

##### Scoping - [x] Scope prior Transformer - [x] Scope decoder Unet - [x] Scope super resolution 64->256 Unet - [x] Scope super resolution 256 ->1024 Unet - [x] Scope...

williamberman

Wrote extended DDIM scheduler that takes pairs of timesteps

1

Re-implemented #1236 a bit differently after a suggestion from @patrickvonplaten . Now the original DDIMScheduler is untouched but a new class is introduced with extra featuers (see below) and DDIMPipeline...

lukovnikov

TypeError: OnnxStableDiffusionPipeline.__init__() missing 1 required positional argument: 'vae_encoder'

12

### Describe the bug Hi, I tried ONNX Runtime for inference. The code is, `from diffusers import StableDiffusionOnnxPipeline pipe = StableDiffusionOnnxPipeline.from_pretrained( "CompVis/stable-diffusion-v1-4", revision="onnx", provider="CPUExecutionProvider", use_auth_token=True, ) prompt = "a photo...

kamalasubha

[Design discussion] Diffusers schedulers/samplers & 2nd Order Schedulers

24

This issue is just used to collect PRs, issues, progress on 2nd order schedulers for Stable Diffusion as the community has asked a lot about this. We'll try to allocate...

patrickvonplaten

add AudioDiffusionPipeline and LatentAudioDiffusionPipeline #1334

2

I have added AudioDiffusionPipeline and LatentAudioDiffusionPipeline which I intend to migrate from https://github.com/teticio/audio-diffusion. I have added them to the main src as opposed to the community pipelines due to the...

teticio

custom pipelines and from_pretrained

4

**What API design would you like to have changed or added to the library? Why?** custom pipelines is available through the master `from_pretrained()` method, but do not work for other...

WASasquatch

mps support under unet_2d_condition.py

3

### Describe the bug mps doesn't support float64. my hacky solution was manually setting timesteps to torch.float32 (line 277) ``` # 1. time timesteps = timestep if not torch.is_tensor(timesteps): #...

d3xtrx

Scale and shift in Resnet block

6

Implemented optional scale and shift (instead of currently only shift) based on timestep embedding, similarly to OpenAI's improved diffusion code. However, it is currently not being used and I would...

lukovnikov