Saving LTN model

[logictensornetworks]

Saving LTN model

[sbadredd]

Hi @maelle101 , thank you for your comment!

At the moment, you need to manually use several tensorflow saving/loading functions for each of your predicates. I imagine that if we have a wrapper like proposed in issue #25 , we could have some saving/loading for the entire theory.

If someday you upload your project somewhere, could you link it here so we can look at what kind of features you expect? Thank you!

[maelle101]

Hello @sbadredd,
Thanks for your answer.
I won't be allowed to share the whole project but I am currently implementing a kind of wrapper for my Ltn architecture.
I may share it, even if it is adapted to my needs, it may be useful for other persons.

Have a good day !

[sbadredd]

Thanks. I appreciate it!