LTU-AI/hdppo-Acrobot-v1

HD-PPO Agent playing Acrobot-v1

This is a trained HD-PPO (Hyperdimensional Proximal Policy Optimization) agent playing Acrobot-v1 using multi-band Fractional Power Encoding (FPE).

Published by LTU-AI.

Pipeline: train multi-band teacher at D=256, prune through 64 → 32 → 16, fine-tune each stage. Published checkpoint is D=16.

Usage

pip install -r requirements.txt
python enjoy.py --weights hdppo-Acrobot-v1/weights.npz --episodes 10
python record_video.py --weights hdppo-Acrobot-v1/weights.npz --output replay.mp4
python enjoy.py --weights LTU-AI/hdppo-Acrobot-v1 --episodes 10

Model files

File	Description
hdppo-Acrobot-v1/weights.npz	Actor/critic/FPE weights (D=16)
replay.mp4	Sample rollout
results.json	Evaluation summary