ali-vilab / AnyDoor

Official implementations for paper: Anydoor: zero-shot object-level image customization
https://ali-vilab.github.io/AnyDoor-Page/
MIT License
4.01k stars 366 forks source link

Train on my dataset #100

Open hithqd opened 4 months ago

hithqd commented 4 months ago

Thank you for your great work! I want to train on my own dataset. What format should this dataset be in?