No batch size? No normalization?

Aminsheykh98 / INCREASE-pytorch

Pytorch version of the INCREASE model.

The Unlicense

3 stars 0 forks source link

No batch size? No normalization? #1

Open mb-Ma opened 4 months ago

mb-Ma commented 4 months ago

I see. The normalization is located in the forward function.

Reset-quick commented 3 months ago

Hello, may I ask if you have now solved the problem that the training set and the validation set in the data set METR-LA become 0 after a few epoches? Can you run out the effect in the paper?

Aminsheykh98 commented 3 months ago

Hello, @Reset-quick

Unfortunately, I haven't resolved the issue. I've been occupied with other projects recently. However, if I resume work on this, I will keep you updated on any progress.

Best regards,