Skip to content
The gym wrapper for MicroRTS for RL research
Python Shell
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
experiments
gym_microrts
.gitignore
README.md
demo.gif
evaluation.bash
hello_world.py
setup.py
test.py

README.md

Gym-MicroRTS

This repo contains the source code for the gym wrapper of MicroRTS authored by Santiago Ontañón.

Get Started

Go to http://microrts.s3-website-us-east-1.amazonaws.com/microrts/artifacts/ to download the latest version of MicroRTS build. For example:

$ mkdir ~/microrts && \
wget -O ~/microrts/microrts.zip http://microrts.s3.amazonaws.com/microrts/artifacts/201911020438.microrts.zip && \
unzip ~/microrts/microrts.zip -d ~/microrts/ && \
rm ~/microrts/microrts.zip
~/microrts

Then clone the repo

$ git clone https://github.com/vwxyzjn/gym-microrts.git && \
cd gym-microrts && \
pip install dacite && \
pip install -e .
$ python3 hello_world.py

And run either the hello_world.py referenced above or the following file

import gym
import gym_microrts

# The prefix `Eval` renders the gmae.
env = gym.make("EvalMicrortsGlobalAgentsProd-v0")
# Alternatively, try headless mode at
# env = gym.make("MicrortsGlobalAgentsProd-v0")
env.reset()
for _ in range(10000):
    env.step(env.action_space.sample())
env.close()

demo.gif

Performances of agents on different map sizes

https://app.wandb.ai/costa-huang/MicrortsRL

You can’t perform that action at this time.