'Nan' loss in train() function?

#1

I was trying to build a neural network to classify point. However, I am not getting ‘nan’ loss during execution of train() function and I am not also getting good accuracy please help.

#2

Hey @HemantKumar, check if you are unsing correct loss function, 'categorical_cross_entropy" for multiple class classification, with ‘softmax’ activation in last layer. For regression problem you must be using ‘mse’ loss with 'linear activation in last layer.

Hope this resolved your doubt.
Plz mark the doubt as resolved in my doubts section. :blush:

#3

reshape Y to (n,) from (n,1)