yxchng
yxchng
Any update on this?
@Jasonbaby Have u solved the problem? I am also training with small batch size 64 with r34 backbone and it is not converging.
@gongliym @minglii1998 have you solved this problem?
any new update on this?
@FruVirus how to convert the fine tuned model to a form load-able by the script? what is the command to run?
1. Is there any good ways to reduce memory usage? I am short on memory and can only train with small batch size. 2. How important is ITC loss? Will...
ok. sorry, i was asking about training the model not inferencing.
@toonn I tried using ueberzug and it says it is only compatible with xterm It is still not working with gnome-terminal.
@rwightman any update on this?
@joliver1981 Have you tried downloading the Kaggle's cc12m? How many images are there?