ltzheng / coinrun

Code for the paper "Quantifying Transfer in Reinforcement Learning"

CoinRun environment for Genie

This is code for the environments used for minimal reproducible case study in Genie.

Install

# Linux
apt-get install mpich build-essential qt5-default pkg-config
# Mac
brew install qt open-mpi pkg-config

git clone https://github.com/openai/coinrun.git
cd coinrun

conda create -n coinrun python=3.6
conda activate coinrun
pip install tensorflow-gpu==1.12.0
pip install -r requirements.txt
pip install -e .

Note that this does not compile the environment, the environment will be compiled when the coinrun package is imported.

Collect 10M transitions

python -m coinrun.random_agent --num-levels 10000 --set-seed 42 --hres