Munchausen-RL icon indicating copy to clipboard operation
Munchausen-RL copied to clipboard

PyTorch implementation of the Munchausen Reinforcement Learning Algorithms M-DQN and M-IQN

Results 2 Munchausen-RL issues
Sort by recently updated
recently updated
newest added

Hello!Excuse me!I used your code in another environment, but I encountered difficulties. Action is a decimal array,, so how to rewrite the " munchausen_addon = log_pi.gather(1, actions)" line of code!...

Should `F.softmax(Q_targets_next, dim=1)` be `F.softmax(Q_targets_next / entropy_tau, dim=1)` instead?