search

lucidrains / deep-daze

Simple command line tool for text to image generation using OpenAI's CLIP and Siren (Implicit neural representation network). Technique was originally created by https://twitter.com/advadnoun
MIT License
4.37k stars 327 forks source link

settings for examples #91

Closed Mearman closed 3 years ago

Mearman commented 3 years ago

sorry didn't know where to put this other than an issue.

if you can remember, would you mind sharing the settings you used to generate the results shown in the readme? I haven't yet been able to develop anything of that quality yet. I get to about 50 images in and it doesn't improve much more.

NotNANtoN commented 3 years ago

Hi!

I created those examples. I used an image size of 256, 44 layers and a batch size of 32 for 8 epochs.

Let me know if that helps!

Mearman commented 3 years ago

Perfect. Thank you.