About training parameters

[Rhysess]

Dear author,

Thanks for sharing the codes of your great work.
In your config.py, you set max epochs to 1000.

May I confirm that you trained for 1000 epochs, then stopped, and used the final weights?
Or you choose the weights of early stage?

[Rhysess]

And may I ask the training time on you GPU?

[haesoochung]

The parameter "max_epoch" is not the actual training epoch. I stopped training after convergence. It took about 4 days for training on RTX 3090. Sorry for my late response!