Ashley Kleynhans

> Yeah I think Exllama by itself is also not working, it seems like some CUDA version issue I am using CUDA 12.1, ExLlamav2 0.0.17 works fine but ExLlamav2 0.0.18...

ExLlamaV2 0.0.19 resolves the issue, I opened a PR to bump ExLLamaV2 to 0.0.19 in the requirements.txt file.

I must have accidentally broken the multiple source images with the new features I added 🙈

Use [RunPod](https://runpod.io/console/gpu-cloud?template=51q837fywe&ref=2xxro4sy), works fine on RunPod and RunPod support is infinitely better than Vast.ai.

Yeah, I am having issues importing my previously saved configs recently, it would be great if the new versions were backwards compatible with previously saved configs.

@bmaltais it happens with the same JSON I shared in #2308

Its working now, thanks @bmaltais

I have the same problem with extremely poor image quality using InstantID, but with A1111 it works perfectly.

> See https://github.com/lllyasviel/stable-diffusion-webui-forge/wiki/ControlNet-Web-API > > Also you probably need to update your forge. Thanks for the wiki link, my forge is already at the latest commit.

Seems I did actually need to update it, but its still not working, now I am getting a different error: ``` *** Error running process: /workspace/stable-diffusion-webui-forge/extensions-builtin/sd_forge_controlnet/scripts/controlnet.py Traceback (most recent call...