cmdr2

Randomize the newly created image name, so that the filename is unique after you right-click and press 'Save'. Right now it always shows 'index.png' or something default. Probably need to...

Hi, This is a minor change for reporting the reason for an Internal Server Error, if an Exception is thrown by predict(). This is really helpful for REST API users...

Upscaling resolution

2

RealESRGAN seems like a decent start

1-click installer and updater. Uses micromamba to install git and conda into a contained environment (if necessary) before running the normal installation script

5

Hi, This PR enables a user to install this project without requiring them to have anaconda or git pre-installed. This reduces unnecessary friction for users who aren't programmers or very...

Allow installing via pip install -e <repo link> - Adds a setup.py

3

Hi, thanks for the excellent library! :) Installing this repo using `pip install -e git+https://github.com/crowsonkb/k-diffusion#egg=k_diffusion` results in a `FileNotFoundError` since it looks for a `setup.py`. I was able to use...

Keep the v1-inference.yaml file, to be able to load SD1 models

1

The SD 2.0 code is capable of loading the older models (SD 1.4), but needs the old v1 config file to load. This PR includes the earlier `v1-inference.yaml` config file,...

It's very useful to use the `img_callback` (added by stability) in the standard PLMS and DDIM `sample()` methods. Unfortunately the DDIM `decode()` (used for img2img) and DPM solvers are missing...

Don't hardcode the device id to 'cuda', instead use what the model is configured to

1

The current implementation hardcodes the device id in the sampler to 'cuda', even though `self.model` knows the correct device. Removing this hardcoding will help support multiple GPUs. Thanks! Credit: @madrang...

Add the SDE variant of DPM-Solver and DPM-Solver++ to DPM Single Step

21

Following up on our conversation at https://github.com/huggingface/diffusers/issues/4167#issuecomment-1648464442 Uses the same approach as https://github.com/huggingface/diffusers/pull/3344/files **Note:** This PR is not ready to merge yet, since it's not clear whether to use `sigma_s0`...

Textual inversion embeddings don't work anymore. Is the plan to deprecate them, or are they work-in-progress?

3

Hi, Apologies if I missed anything. I noticed that the codepaths for loading embeddings was commented out in the "Free WebUI from its Prison" commit - https://github.com/lllyasviel/stable-diffusion-webui-forge/commit/bccf9fb23a2cbc9d42091a856b3d35b2c4414d51 This results in...