Pytorch Reinforcement Learning

This repository contains the code for policy gradient algorithm incorporating with credit assignment mechanism.

Install Dependencies

1. Install Pytorch

pip install torch torchvision

2. install Tensorflow 2

pip install tensorflow=2.2

or

pip install tensorflow-gpu=2.2

3. Install OpenAI baseline (Tensorflow 2 version)

git clone https://github.com/openai/baselines.git -b tf2 && \
cd baselines && \
pip install -e .

Note: I haven't tested the code on Tensorflow 1 yet but it should work as well.

4. Install gym

pip install 'gym[atari]'

5. Install Park Platform. I modified the platform slightly to make it compatible with OpenAI's baseline.

git clone https://github.com/lehduong/park -b openai_baseline &&\
cd park && \
pip install -e .

Run experiments

python main.py --algo a2c --env-name PongNoFrameskip-v4

Acknowledgement

The started code is based on ikostrikov's repository