README.md

metadata

license: apache-2.0
tags:
  - reinforcement-learning
  - ml-agents
  - ppo
  - unity

PPO Agent playing Huggy 🐻

This repository contains a trained PPO agent playing Huggy, using the Unity ML-Agents toolkit.

Training Details

• Algorithm: PPO
• Framework: Unity ML-Agents
• Observation space: Vector observations
• Action space: Discrete / Continuous (adjust if needed)

Files

• checkpoint.pt: Trained model weights
• configuration.yaml: Training configuration
• events.out.tfevents.*: TensorBoard logs

Usage (with ML-Agents)

Resume training

mlagents-learn configuration.yaml --run-id=Huggy --resume