Hi, I have the following problem - when running a tensorflow model (nn) I have two logs printed out through the model (with tensorboard), one for the mse/loss and one for the val_loss (which are the same). The reports of the mse/loss returned from my tf model don't align with the scalars reported (but for the val_loss the reports are consistent)
The mse/loss report is via :
model.compile(optimizer=adam, loss='mse', metrics=['mse'])