Hi John,
I actually went back and added the tau update to DDQN after posting the code, but hadn’t done the same for DRQN. The tau update would likely improve stability for DRQN as well.
Hi John,
I actually went back and added the tau update to DDQN after posting the code, but hadn’t done the same for DRQN. The tau update would likely improve stability for DRQN as well.
PhD. Interests include Deep (Reinforcement) Learning, Computational Neuroscience, and Phenomenology.
PhD. Interests include Deep (Reinforcement) Learning, Computational Neuroscience, and Phenomenology.