Williams-Hao / Chapter13-Learning-to-Run

Chapter 13 Learning to Run in book Deep Reinforcement Learning: code example of solving NIPS 2017: Learning to Run challenge with paralleled Soft Actor-Critic (SAC) algorithm.

Chapter 13: Learning to Run (Project 1)

Description:

Example of solving NIPS 2017: Learning to Run challenge with paralleled Soft Actor-Critic (SAC) algorithm.

Dependencies:

• osim-rl: install the osim-rl environment following here or using the docker (docker pull stanfordnmbl/opensim-rl)
• PyTorch

The environment dependencies can be installed as follows with a new conda env (opensim-rl):

# 1. Create a conda environment with the OpenSim package.
conda create -n opensim-rl -c kidzik opensim python=3.6.1

# 2. Activate the conda environment we just created.
# On windows, run:
activate opensim-rl

# On Linux/OSX, run:
source activate opensim-rl

# 3. Install our python reinforcement learning environment.
conda install -c conda-forge lapack git
pip install osim-rl

# 4. Git clone the Nips2017 environment and our solution.
git clone https://github.com/deep-reinforcement-learning-book/Project1-RL-for-Learning-to-Run

If installed correctly, run following script should start the env:

from osim.env import RunEnv
env = RunEnv(visualize=True)
observation = env.reset(difficulty = 0)
for i in range(200):
    observation, reward, done, info = env.step(env.action_space.sample())

Contents:

• osim/: the original version of osim-rl for NIPS 2017: Learning to Run challenge, osim-rl environments have been updated and no longer provide 2017 version through direct package installation;
• figures/: figures for displaying;
• model/: models after training;
• sac_learn.py: pralleled Soft Actor-Critic algorithm for solving NIPS 2017: Learning to Run task;
• reward_log.npy: log of episode reward during training;
• plot.ipynb: displaying the learning curves.

Usage:

1. Run $ python sac_learn.py --train for training the policy

2. Run $ python sac_learn.py --test for testing the trained policy, remember to change the trained_model_path for testing your own model, which is default to be the trained model we provided.

3. The training process will provide a reward_log.npy file for recording the reward value during training, which can be displayed with $ jupyter notebook in a new terminal, choose plot.ipynband Shift+Enter, as follows:

Authors:

Zihan Ding, Yanhua Huang

Citing:

@book{deepRL-2020,
 title={Deep Reinforcement Learning: Fundamentals, Research, and Applications},
 editor={Hao Dong, Zihan Ding, Shanghang Zhang},
 author={Hao Dong, Zihan Ding, Shanghang Zhang, Hang Yuan, Hongming Zhang, Jingqing Zhang, Yanhua Huang, Tianyang Yu, Huaqing Zhang, Ruitong Huang},
 publisher={Springer Nature},
 note={\url{http://www.deepreinforcementlearningbook.org}},
 year={2020}
}

or

@misc{DeepReinforcementLearning-Chapter13-LearningtoRun,
  author = {Zihan Ding, Yanhua Huang},
  title = {Chapter13-LearningtoRun},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/deep-reinforcement-learning-book/Chapter13-Learning-to-Run}},
}