Policy-Gradient-Methods
Policy-Gradient-Methods copied to clipboard
Is your DDPG implement missing adding noise?
I find you write: self.noise = OUNoise.... but you didn't add the noise to the action?
so I add the following line to line48 in ddpg/ddpg.py: action = self.noise.get_action(action) # add by xiaotong 2020 0409
If I made something wrong, please tell me, thanks.