Hacker News new | past | comments | ask | show | jobs | submit login

It's interesting to see that Latent Diffusion + LAION 400M does much better with text, it's actually a challenge sometimes to get it to not write the parts of the prompt meant to guide the style on the resulting image.

Try it yourself here: https://colab.research.google.com/github/multimodalart/laten...

EDIT: Here's 36 uncurated renders for "Hacker News - Dang's Dance. R&B Album Artwork" https://johan-nordberg.com/tmp/hn-laion400m.png




First link colab doesn't work, output_path variable not declared when running the last block even if first one, censed to declare it, is run.


Try Runtime -> Run All

Or use the huggingface version, that's a simple web ui: https://huggingface.co/spaces/multimodalart/latentdiffusion


by default nsfw_threshold=0.5

nsfw_threshold=1 for no filtering


Nice. Now, let's have it generate some FLAC files.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: