HI @robertl
Settings of the entire model at a glance??
I’ve just set up something to dump the “key” parameters of my TF model in python to aid in reproducibility - and there are a lot (especially when basic ideas like early stopping involve additional parameters e.g. for early stopping - baseline or delta, patience, etc.)
[NB I’m doing this so that I can do offline hyperparameter “tuning” based on more history than I might think of in using the TF hyperparams thing and/or when there isn’t time to let the machine run all the variations]
It would be great if one could summarise a whole model “at a glance” but even if you could, I doubt the implementation would be quick or easy.
I admire the ambition though!