Skip to content

A simple stochastic OpenAI environment for training RL agents

Notifications You must be signed in to change notification settings

lmerchante/hearthstone-gym

 
 

Repository files navigation

This repository contains a PIP package which is an OpenAI environment for simulating an environment in which Hearthstone is played.

Installation

Install the OpenAI gym.

Install Fireplace Hearthstone Simulator

git clone https://github.com/jleclanche/fireplace
cd fireplace
pip install .

Then install this package via

git clone https://github.com/SethKitchen/hearthstone-gym
cd hearthstone-gym
python -m pip install -e .

Usage

import gym
import gym_hearthstone

env = gym.make('Hearthstone-v0')
actions=env.get_possible_actions()
env.step(actions[0])

The Environment

Uses Fireplace and random MAGE/WARRIOR decks -- updates to expand

With Help From

https://github.com/MartinThoma/banana-gym

https://github.com/albertwujj/HearthEnv

About

A simple stochastic OpenAI environment for training RL agents

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 99.6%
  • Other 0.4%