Pedro Cuenca

There are some [specialist in-painting models](https://huggingface.co/runwayml/stable-diffusion-inpainting), but by and large most stable diffusion models are capable of many of those tasks @Narsil outlined, at least to some extent. Perhaps we...

Yes, of course! My question was more about what models should be tagged with these new widgets vs just with the text-to-image task. If I understand it correctly, the widget...

I think this was resolved in #1120.

Fixed by #1668 (except keeping the last `n` checkpoints, to be adapted from https://github.com/huggingface/accelerate/issues/914).

Update: this works better in PyTorch 2 but still fails occasionally. For example, the following code works for `num_samples = 3`, but fails for `num_samples = 2`. (Yes, `2` fails...

Hi @jverkoey! I recently proposed [this PR](https://github.com/apple/ml-stable-diffusion/pull/59), which does two things: - Convert `Scheduler` into a protocol. - Implement a new scheduler (DPM-Solver++) using that protocol. If it is accepted,...

Hi @allo-! Yes, #1687 implements checkpointing for training scripts, and those checkpoints can be used for inference using a process like this: https://huggingface.co/docs/diffusers/main/en/training/dreambooth#performing-inference-using-a-saved-checkpoint I'm sorry, but I still don't fully...

@mlmcgoogan @patrickvonplaten that's been my experience too. @antoche I've seen similar improvements on 3090 as those reported by @mlmcgoogan, and we have also tested on V100, A100 and others. We...

I'm curious about this. Are those embeddings related to textual inversion training, or something else? Would you happen to have any pointers?

@atiorh @littleowl I think `DPMSolverMultistepScheduler` should also work. I just double-checked in `diffusers` and it did. I'll try to debug what might be going on.