Implement some features that can make TI training 6x faster
Cafeai from 東方Project AI used some cool tricks to make TI training 6x faster compared to auto webui. I don't understand them so I don't know if they has been implemented in your repo. If not then can you implement them?
https://github.com/cafeai/stable-textual-inversion-cafe
Thank you for letting me know!
I briefly checked the repo, but unfortunately I cannot find which part contributed to the speed improvement. If you have any information for that, please share.
I think extended_mode is the difference which contributed to the speed improvement
https://github.com/nicolai256/Stable-textual-inversion_win/commit/74c6646acd3876e9493c505ec3ea31530642745b#diff-426258e702a37e401644429f61e4e54c688707509e5502ae2cc5f87e51dec1b7