You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, great work! I have a question regarding the fine-tuning process. Since the model adopts a Decomposed Dual-Branch Diffusion architecture—comprising BrushNet and the original UNet—should I fine-tune BrushNet, UNet, or both branches during training?