Validation loss

Open stavyadatta opened this issue 3 years ago • 0 comments

Hello, I am new to training facial recognition models and would greatly appreciate any insights you may have. Can you explain how to calculate validation loss for a facial recognition model. I do realise that the head is different for training and validation. Is there a way to calculate validation loss? Additionally, could you provide further information on line 265 in the train_tripletloss.py file, as it references validation loss but I do not see any calculation of it. Thank you.

Jan 11 '23 13:01 stavyadatta