Hi
When I’m running the Keras Learner node, I feel the need to record some metrics from the training process. This way, I can use them together with the settings of the Keras Learner to investigate the effect of these settings on the result of my predictions.
Currently, I can already record many settings via the Flow Variable tab. However, what is not yet recorded, and what I do need, are:
In which epoch the training stopped.
What the current value (accuracy / loss) of the Training Data is.
And what the current value (accuracy / loss) of the Validation Data is.
gr. Hans