Chenxi

Hello again @xinntao :), Thank your for your previous feedback on the original PR, it has been updated accordingly. Also we are very happy that you have claimed the [page](https://replicate.ai/xinntao/gfpgan),...

> # No Dockerfile > > > > * I don't see anything in this PR :) > > * To me, Docker environment is when I can do `docker-compose...

@simonmaris hi you can find the files here https://github.com/chenxwh/dalle-mini :)

> Fantastic, thank you! Quick question before we do the work twice: were you planing to add [yielding of progressive output](https://github.com/replicate/cog/blob/main/docs/python.md#progressive-output) or are these just done for you currently? Go...

> This leaks the text2img model checkpoint ? the checkpoint is publicly available from the repo

> > ? the checkpoint is publicly available from the repo > > I'm not sure what you're talking about, all references in the repo link to Stability's internal cluster?...

Aware of latent-diffusion. Maybe it should be stated in the readme, similarly someone asked [here](https://github.com/pesser/stable-diffusion/issues/1#issuecomment-1182610546) but still using this way

@metaphorz Here is Replicate demo that is backed with A100 if you are interested :) https://replicate.com/thudm/cogview2

Hi, I have checked the CUDA usage locally, it does seem some memory persists after a `predict()` run, I have pushed to the website a new version freeing about 10G...

Hi, you can create as many demo pages as you like on Replicate. The document here https://github.com/replicate/cog and the demonstration in this PR should help you with deploying and updating...