Monitor validation loss

I’m training an intent classifier using my custom dataset.
I’m using tensorboard for monitoring the training process, but I don´t have access to the validation loss, it only plots the validation accuracy(in blue):


Is there any way to tell the trainer to include the validation loss in the logs?

Hello!
Thank you for the question. To monitor loss (or any other metric you want), you should add it to config[“train”][“metrics”].

Hi!
Thanks for the fast reply.
Do you mean that I must add “val_loss” as a metric in the config[“train”][“metrics”]?