Open lofangmeng opened 1 month ago
Hello author,
I would like to ask about the image generation part. When using coadapter, have you trained the coadapter, or did you use it directly? Also, is the generated fMRI clip space embedding [768] directly inputted to replace the prompt?
Hello author,
I would like to ask about the image generation part. When using coadapter, have you trained the coadapter, or did you use it directly? Also, is the generated fMRI clip space embedding [768] directly inputted to replace the prompt?