Hello AI-Friends,

I am using a callback function to calculate train and test error after each epoch end by calling model.evaluate() . However if I compare my calculated train loss from model.evaluate(x_train, y_train) it is different from the loss saved in history.history['loss']. The calculated ...