bad result on Antmaze enviornment

BY571 / Implicit-Q-Learning

PyTorch implementation of the implicit Q-learning algorithm (IQL)

42 stars 4 forks source link

bad result on Antmaze enviornment #2

Open dljzx opened 2 years ago

dljzx commented 2 years ago

It works well on mujoco environments, but not on antmze environment .It did not work even if I changed the parameters according to the paper(expectile=0.9, temperature=10). Can you help me please?

BY571 commented 2 years ago

I'm not sure what the problem is but I will have a closer look once I have time. Could be that its caused due to the cosine schedule for the actor learning rate which I didn't add yet.