ppo-Huggy / README.md
Zybg's picture
Add model card for Huggy Space
e064201 verified
metadata
library_name: mlagents
tags:
  - unity
  - ppo
  - huggingface-deep-rl-course
  - Huggy

🐶 PPO-Trained Huggy Agent

This model was trained using Unity ML-Agents (PPO) to teach Huggy the Dog to fetch a stick.

➡️ Play with it here:
👉 https://huggingface.co/spaces/ThomasSimonini/Huggy

🧠 Training Details

  • Algorithm: PPO
  • Environment: Huggy (fetch-the-stick dog game)
  • Framework: Unity ML-Agents v1.2.0
  • Reward Design: +orientation bonus −time penalty −rotation penalty
  • Max Steps: 2 × 10⁶
  • Checkpoint interval: 200 000

You can replay this model in the Huggy Space by entering: