I am trying to do the following:
- Install the OpenAI baseline algorithms from the following GitHub repository: github.com/openai/baselines by following the instructions in the readme file.
- Train an agent on the pendulum task by running DDPG: "python -m baselines.ddpg.main --env-id Pendulum-
Once I do that, I need to modify the DDPG code in baselines/ddpg/ to share the weights between the first layer of the actor network and the first layer of the critic network.
Finally, I need to produce a graph where the y-axis is the average reward of each episode obtained by DDPG in the Pendulum-v0 environment and the x-axis is the # of episodes up to a maximum of 500 episodes.
I am unsure how to modify the DDPG code, as I am new to tensorflow. Can any one help me out here, or guide me as to how I can do that?