tf_slim is very much idiomatically TF1, to the point that I'm not sure it works when we are executing eagerly/utilizing tf.function. We should replace usages of it (mostly continuous evaluation in main_function.py) with a custom loop utilizing new TF2 APIs.
tf_slim is very much idiomatically TF1, to the point that I'm not sure it works when we are executing eagerly/utilizing tf.function. We should replace usages of it (mostly continuous evaluation in
main_function.py) with a custom loop utilizing new TF2 APIs.