sled-group / InfEdit

[CVPR 2024] Official implementation of CVPR 2024 paper: "Inversion-Free Image Editing with Natural Language"
https://sled-group.github.io/InfEdit/
Other
242 stars 7 forks source link

how to adapt lora to lcm? #9

Closed leonardodora closed 3 months ago

leonardodora commented 5 months ago

Amazing work! But I wonder is there possible to add lora to the base lcm ?

SihanXU commented 4 months ago

Yes. We use diffusers pipeline as our code base, so I think you can do whatever you want like add loras to the model, or use difference checkpoints.

leonardodora commented 4 months ago

Yes. We use diffusers pipeline as our code base, so I think you can do whatever you want like add loras to the model, or use difference checkpoints.

Thanks! That means I need to train a lora from LCM_Dreamshaper_v7 then I can merge the lora to LCM_Dreamshaper_v7, Or I can directly use lora trained from sd1.5?

SihanXU commented 4 months ago

I cannot give your the answer as well. Just play with your loras!