Hi, thanks for sharing this useful work! I'm reading the program and there was something wrong when I debug 'main.py' (train). The problem is as following:
for ema_decay in args.model_ema_decay:
TypeError: 'float' object is not iterable
'args.model_ema_decay' was set 0.99996 as guideline.
Could you give some suggestions on the problem? Thanks a lot.
you can directly change the default args.model_ema_decay from 0.99996 to [0.99996], this for loop is used to train with multiple ema decay numbers simultaneously.
Hi, thanks for sharing this useful work! I'm reading the program and there was something wrong when I debug 'main.py' (train). The problem is as following:
'args.model_ema_decay' was set 0.99996 as guideline. Could you give some suggestions on the problem? Thanks a lot.
Yue.