Number of nodes in a hidden layer are different from the DDPG paper

stevenpjg / ddpg-aigym

Continuous control with deep reinforcement learning - Deep Deterministic Policy Gradient (DDPG) algorithm implemented in OpenAI Gym environments

MIT License

275 stars 74 forks source link

Number of nodes in a hidden layer are different from the DDPG paper #3

Closed sarvghotra closed 8 years ago

sarvghotra commented 8 years ago

https://github.com/stevenpjg/ddpg-aigym/blob/master/critic_net.py#L58 According to the paper this should be 300.

stevenpjg commented 8 years ago

Fixed! 2nd layer has been changed to 300 as mentioned in the ddpg-paper. The output of the actor-net is always near the extreme i.e) either near -3 or +3. Still trying to figure out the error. Feel free to contribute to this code.