search

jsphon / reinforcement_learning

Python Package For Reinforcement Learning
0 stars 0 forks source link

tf loop #42

Open jsphon opened 6 years ago

jsphon commented 6 years ago

Our line world example works for single training ops, but not for train_loops.

Fix it for train loops

jsphon commented 6 years ago

I'm not too sure where to begin as the error message is pretty cryptic.

Perhaps I could see if test_train_loop fails if I use a one_hot input on it. If that works, gradually addd the other features in our system.py example, to the test case, until we get a failure.