Char 05 DDPG missing exploration noise

sweetice / Deep-reinforcement-learning-with-pytorch

PyTorch implementation of DQN, AC, ACER, A2C, A3C, PG, DDPG, TRPO, PPO, SAC, TD3 and ....

MIT License

3.88k stars 844 forks source link

Char 05 DDPG missing exploration noise #3

Closed tomas-gajarsky closed 5 years ago

tomas-gajarsky commented 5 years ago

The authors of DDPG are adding noise to the action (Section 7 - Experimental details), in order to explore more options. This feature is missing in DDPG.py script.

sweetice commented 5 years ago

Thanks for your issue. This bug has been fixed.

# issue 3 add noise to action
action = (action + np.random.normal(0, args.exploration_noise, size=env.action_space.shape[0])).clip(env.action_space.low, env.action_space.high)

: )