Distributed TensorFlow implementation of Asynchronous Methods for Deep Reinforcement Learning.
- Python 2.7 or Python 3.3+
- gym
- tqdm
- OpenCV2 or Scipy
- TensorFlow
First, install prerequisites with:
$ pip install tqdm gym[all]
To train a model with CPU:
$ ps_num=1 worker_num=1 start_port=2222 ./run.sh
(in progress)
MIT License.