Closed marksverdhei closed 3 years ago
Temporary solution: suppress the excessive warnings produced by tensorflow, as I'm pretty sure they do not have any actual impact on our model. Might train a new model with more epochs or parameters after we've gotten a validation framework down. Hopefully, it will not produce these warnings. The solution forces us to make an exception regarding pep8, as addressed in tensorflow issue #31870. We have to set the environment variable before importing keras
LGTM! I'll still keep the issue open though in case you find a permanent solution or decide to close the issue yourself.
A bunch of warnings occur when loading the keras model. Not sure why, but it should either be resolved or suppressed