Adam Letts

Thanks for the PR! It looks all good, and that it'll work. I'll see about picking the version number, etc.

I think you may have installed that one locally ( ? ). I remember seeing that once when I was initially integrating AdaBins in my local setup. I think `infer`...

Actually, I think I see what you're saying now. You may be saying that you have `infer` for a real reason - and that you want to be able to...

It'd be easiest for me to adjust DD's installation process for it.. to make it rename infer.py to adabins_infer.py or something. It's a hack, but it wouldn't be the first...

Currently, Disco Diffusion doesn't support this. In theory there are ways to do this.. but it hasn't been implemented. Its history of being primarily used via Google Colab (and by...

It's easy to make it run on a different GPU, an it's something that I do sometimes. You can search for the "cuda:0" string in disco.py and change it to...

For maximum cohesiveness, it can be a good idea to match the output resolution to the diffusion model's trained image size. So in this case, you'd want to experiment with...

I've got some changes in progress so I won't merge it right now.. but I think the following awkward, untested code may work: ``` if resume_run: if run_to_resume == 'latest':...

This looks cool. The main consideration I have at this point is whether or not `colab-convert` can convert this .ipynb->.py->.ipynb without losing the data. @MSFTserver : You'll likely be interested...

There are at least a couple OpenAI papers that are directly involved. These are good starting points.. and there are good links throughout and in the citations. CLIP - Learning...