Traing loss

[XyFighting]

Hi, thanks for your amazing work!

I've encountered an issue with NaN losses when using MambaVision to train on CIFAR100 dataset. Could you suggest solutions to solve it?

[ahatamiz]

Hi @XyFighting , the best recommendation is to lower the learning rate. The settings used in ImageNet experiment may not be optimal for other datasets.

[XyFighting]

Hi, thanks for your reply!
I have decreased the learning rate. However, when the training process proceeds after dozens of epochs, there are still NAN losses. Could you further suggest solutions to solve it?