LSTM State is missing in generating the action #2
Description
I think when actor is generating the next state, it fails to use the initial state from last cell, as in this line trainer.py
{{ message }}
I think when actor is generating the next state, it fails to use the initial state from last cell, as in this line trainer.py