AnimeRun icon indicating copy to clipboard operation
AnimeRun copied to clipboard

Published 20 hours ago • verified publisher icon lisiyao21

About the pre-trained weight of AnT model.

Open JCVBoss opened this issue 2 years ago • 4 comments

Thank you for such an inspiring work. Noticed that your code uses a re-implementation of AnT to replace the fragment matching in the original SGM Flow, I have some own data, the original SGM Flow doesn't work very well on it. If you can provide the pre-trained weights of AnT model, which would be very helpful.

JCVBoss avatar Nov 17 '22 03:11 JCVBoss

Thank you so much for your interest! Sorry I've been a mess recently. I think I can clear the code and release the weights before the end of this month.

lisiyao21 avatar Nov 17 '22 16:11 lisiyao21

hi guys! author of AnT here. thank you for this work! it's very exciting. happy to answer any questions about AnT here

evancasey avatar Dec 24 '22 20:12 evancasey

@JCVBoss Sorry I got Covid19 so I delayed. I'm trying to work on it soon...

lisiyao21 avatar Dec 25 '22 16:12 lisiyao21

hi guys! author of AnT here. thank you for this work! it's very exciting. happy to answer any questions about AnT here

Oh Hi!! Very glad to see you here!! AnT is very cool. I hope you could find our data useful :D

lisiyao21 avatar Dec 25 '22 16:12 lisiyao21