/docker-rl-gym

Primary LanguageDockerfile

Docker - RL & OpenAI Gym

Description:

A Docker environment for reinforcement learning in Python 3 including the OpenAI Gym toolkit

Includes:

1. Basics: NumPy, Pandas, Scipy, Jupyter, Matplotlib

2. Deep Learning: TensorFlow, Keras

3. Reinforcement Learning:Keras-RL, baselines, TensorForce

4. Environments:AI Gym

5. Others: ipywidgets, h5py

Rendering on Jupyter notebook

The virtual frame buffer allows the video from the gym environments to be rendered on jupyter notebooks. Simple example with Breakout:

import gym
from IPython import display
import matplotlib.pyplot as plt
%matplotlib inline

env = gym.make('Breakout-v0')
env.reset()
for _ in range(1000):
    plt.imshow(env.render(mode='rgb_array'))
    display.clear_output(wait=True)
    display.display(plt.gcf())
    env.step(env.action_space.sample())