maurorisonho/ppo-BreakoutNoFrameskip-v4

PPO Agent Playing Atari BreakoutNoFrameskip-v4

This is an advanced reinforcement learning agent trained with PPO and Deep Convolutional Networks (CnnPolicy) to play classic Atari Breakout, representing the bonus mastery challenge of the Hugging Face Deep Reinforcement Learning Course (Unit 3).

🚀 Model Details

• Environment: ALE/BreakoutNoFrameskip-v4
• Algorithm: PPO with 4-Frame Stacking and Nature CNN Backbone
• Mean Score: 412.0 points
• Files: Policy weights (.zip), video replay (replay.mp4)