TencentQQGYLab / ELLA

ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment
https://ella-diffusion.github.io/
Apache License 2.0
1.1k stars 57 forks source link

Prompt syntax (eg. Compel) & ELLA #50

Open andupotorac opened 4 months ago

andupotorac commented 4 months ago

Hi. What happens if we are to use InvokeAI-like syntax in a prompt, and pass that to ELLA? Will it know how to pass it further to maintain the right weights for specific words?

andupotorac commented 4 months ago

Is it fair to assume this approach used for CLIP with TIs will also work here with weighed tokens? https://github.com/TencentQQGYLab/ELLA?tab=readme-ov-file#3-ellaclip-for-community-models