/dqn_implementation

Implementation of DQN as specified in Human Level Control Through Deep Reinforcement Learning

Primary LanguageJupyter Notebook

DQN Implementation

Implementation DQN as specified in Human Level Control Through Deep Reinforcement Learning. Knowledge of the paper is expected.

Installation

  1. Create a virtual environment (venv, pipenv, conda, ...) with Python 3 (Python 3.7.4 recommended) and pip
  2. Execute pip install -r requirements.txt
  3. Navigate into the root directory (containing setup.py) and install package by pip install -e .
  4. Make sure you have CUDA installed (essentially any version compatible with pytorch, see below)
  5. Install pytorch to match your CUDA version (project was built for pytorch 1.2, so keep this in mind when deciding what pytorch-CUDA combination to install)

Current state

Work in progress. See Github project of this repository for details.