incandescent.

I didn't end up producing as much today as I expected, fell down a rabbit hole. I wanted to be able to run LLM in comfy, easy enough after trying numerous broken extensions but vision models (they can see) need a python install. I keep multiple pythons, I ended up installing and reinstalling packages on all of them, followed every guide, eventually found comfy uses an independent python that I apparently wasn't connecting to. Got it working but still need gpu acceleration. Will it be useful? probably not but it's cool XDD