psychedelicious

IMO an extra VAE encode and some additional support code for inpainting is an acceptable tradeoff for no possibility of artifacts. I'm imagining a scenario where you inpaint and it's...

I believe the root issue is the lack of proper RAM and VRAM management of these models. #6132 will resolve that.

Did it work previously? If so, what was the last app version where it worked? This line is interesting: ``` /home/invokeuser/venv/lib/python3.10/site-packages/diffusers/image_processor.py:90: RuntimeWarning: invalid value encountered in cast images = (images...

Ok. I believe this issue is related to the GPU and its supported precision settings. Let's try forcing fp32 precision for the whole app. In the `invokeai.yaml` file, add a...

How about a larger text box where you can put a model on each line (or separate multiples with commas)?

Sorry, this isn't particularly high on the list of features. There's no timeline for when it might be added, but we'd welcome a PR adding it. Probably wouldn't be too...

The handling of tiled decode hasn't changed in some time - several months. This functionality is handled wholly by `diffusers`, and it appears their implementation also hasn't changed in months....

Thanks for those examples. It's still very noticeable. I think we need to test this with diffusers (i.e. via separate script, not within invoke) to confirm where the problem is.

This was inadvertently broken in #5991. The NSFW checker model isn't downloaded unless it is requested, but if it isn't downloaded yet, the app reports it is unavailable - catch...

Leave this open and you'll get notifications when it's fixed (or if we end up removing it). Actually, I already fixed it but my strategy had some issues. You can...