search

mit-han-lab / fastcomposer

[IJCV] FastComposer: Tuning-Free Multi-Subject Image Generation with Localized Attention
https://fastcomposer.mit.edu
MIT License
670 stars 38 forks source link

is it possible to use fastcomposer with SD XL ? #17

Open SlZeroth opened 1 year ago

SlZeroth commented 1 year ago

I want to try fastcomposer code with SD XL for better result how to use it with XL?

ucasligang commented 1 year ago

Same question!

SlZeroth commented 1 year ago

@ucasligang do you have any plan to implement fastcomposer support the XL? I think move this repo code to support XL... with train again

ucasligang commented 1 year ago

@ucasligang do you have any plan to implement fastcomposer support the XL? I think move this repo code to support XL... with train again

We also plan to implement fastcomposer support for the XL.

SlZeroth commented 1 year ago

@ucasligang plan to pull request to this repo?

SlZeroth commented 1 year ago

@ucasligang hello? any update?

chenjun0210 commented 1 year ago

Same question!

liguanglin commented 10 months ago

cant wait to see the xl version, any update?

tanghengjian commented 6 months ago

hi, https://github.com/tanghengjian/fastcomposer/commits/sdxl/ FYI @SlZeroth @chenjun0210 @liguanglin @songhan
i have tried to modify some codes in sdxl, but the inference result is pure black image, may some issue inside.