diffusers

diffusers copied to clipboard

Add 2nd order heun scheduler

2

This PR shows how we can integrate Heun's scheduler into our current framework **without** any changes to the pipelineis. We simply stretch the timesteps and sigmas => I honestly don't...

patrickvonplaten

K-Diffusion + Stable Diffusion Pipeline

3

### Model/Pipeline/Scheduler description It would be really nice to have a pipeline that integrates well with k-diffusion so that all schedulers be used out of the box will all checkpoints...

patrickvonplaten

Improve DDIM sampling pipeline (support for predict x0 and uneven jumps)

1

* support for predict_epsilon = False in DDIM sampler * changed timestep selection such that it is more uniform if number of sampling steps doesn't cleanly divide number of training...

lukovnikov

Unused parameters in scheduler `init`'s

12

**What API design would you like to have changed or added to the library? Why?** I see some of the [schedulers](https://github.com/huggingface/diffusers/blob/7bd50cabafc60bf45ebbe1957b125d3f4c758ba8/src/diffusers/schedulers/scheduling_lms_discrete.py#L34) had these parameters added to init ``` trained_betas: Optional[np.ndarray]...

natolambert

Same seed produces different results

13

### Describe the bug Basically is you set the scheduler to EulerAncestralDiscreteScheduler and the custom pipeline to lpw_stable_diffusion you will get different images when you generate ### Reproduction Here is...

thisislance98

[Proposal] Support saving to safetensors

1

Follow-up of https://github.com/huggingface/diffusers/pull/1357, and mimics Transformers https://github.com/huggingface/transformers/pull/20321/files#diff-82b93b530be62e40679876a764438660dedcd9cc9e33c2374ed21b14ebef5dba

MatthieuBizien

Do not use torch.long in mps

1

Fixes #1056. Another option is to unconditionally use `torch.float32` in all platforms (both `int` and `float` are accepted as inputs), what do you think?

pcuenca

DummyObjects are impeding PyCharm

2

I'm subclassing StableDiffusionPipeline (because it seems like that's the intended way to make a DiffusionPipeline that is still able to take advantage of StableDiffusionPipeline's methods to enable attention slicing, decode...

keturn

[refactor] Making the xformers mem-efficient attention activation recursive

3

- move the enable/disable call to being part of the base DiffusionPipeline (removes a bunch of duplicates) - make the call recursive across all the modules in the model graph,...

blefaudeux

Compute embedding distances with torch.cdist

3

20Go -> 16Go Ram use for some workloads, same speed (you don´t have to materialize intermediates with torch.cdist)

blefaudeux