Dqn SpaceInvadersNoFrameskip V4

🚀 Stable-Baselines3 DQN Agent for SpaceInvadersNoFrameskip-v4

This project presents a trained DQN agent designed to play the SpaceInvadersNoFrameskip-v4 game. It leverages the stable-baselines3 library and the RL Zoo to achieve efficient reinforcement learning training and deployment.

📋 Model Information

🚀 Quick Start

Prerequisites

• RL Zoo: https://github.com/DLR-RM/rl-baselines3-zoo
• SB3: https://github.com/DLR-RM/stable-baselines3
• SB3 Contrib: https://github.com/Stable-Baselines-Team/stable-baselines3-contrib

Installation

Install the RL Zoo (with SB3 and SB3-Contrib):

pip install rl_zoo3

Usage

Download the model and save it into the logs/ folder:

python -m rl_zoo3.load_from_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -orga ThomasSimonini -f logs/
python -m rl_zoo3.enjoy --algo dqn --env SpaceInvadersNoFrameskip-v4  -f logs/

If you installed the RL Zoo3 via pip (pip install rl_zoo3), you can run the above commands from anywhere.

💻 Usage Examples

Basic Usage

# Download model and save it into the logs/ folder
python -m rl_zoo3.load_from_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -orga ThomasSimonini -f logs/
python -m rl_zoo3.enjoy --algo dqn --env SpaceInvadersNoFrameskip-v4  -f logs/

Advanced Usage

If you installed the RL Zoo3 via pip (pip install rl_zoo3), from anywhere you can do:

python -m rl_zoo3.load_from_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -orga ThomasSimonini -f logs/
python -m rl_zoo3.enjoy --algo dqn --env SpaceInvadersNoFrameskip-v4  -f logs/

📦 Training

Training with the RL Zoo

python -m rl_zoo3.train --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/
# Upload the model and generate video (when possible)
python -m rl_zoo3.push_to_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/ -orga ThomasSimonini

🔧 Technical Details

Hyperparameters

OrderedDict([('batch_size', 32),
             ('buffer_size', 100000),
             ('env_wrapper',
              ['stable_baselines3.common.atari_wrappers.AtariWrapper']),
             ('exploration_final_eps', 0.01),
             ('exploration_fraction', 0.1),
             ('frame_stack', 4),
             ('gradient_steps', 1),
             ('learning_rate', 0.0001),
             ('learning_starts', 100000),
             ('n_timesteps', 10000000.0),
             ('optimize_memory_usage', False),
             ('policy', 'CnnPolicy'),
             ('target_update_interval', 1000),
             ('train_freq', 4),
             ('normalize', False)])

Environment Arguments

{'render_mode': 'rgb_array'}