/neural-slimevolley

Neuroevolution for slime volleyball :D

Primary LanguagePython

NEAT neural slime volleyball RainbowBlobCat

Evolves an agent through tournament self-play using NEAT-Python on SlimeVolley task.

  1. set NEAT parameters in config-feedforward
  2. set training parameters in train.py
  3. run python3 train.py

See output logs in log/neat-slimevolley for final model .gif, saved checkpoints, and network graphs.

Uses feature engineering, and hit reward annealing to speed up convergence.

final agent + example evolved topology

final_agent topology