Training and validation loss greater than 1

[akshayk0406]

Hi,

I am sequential model for 3 class classification. When I am training the classifier i am seeing that training and validation loss are greater than 1. How is this possible. Can somebody please explain me ?

[joelthchao]

If you use categorical cross entropy (formula here), it's possible to have loss greater than 1.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed after 30 days if no further activity occurs, but feel free to re-open a closed issue if needed.