Back to browse
GitHub Repository
0 starsPython

RL Environment for Sheep Hearding

by blackboattech·Jul 18, 2026·1 point·0 comments

AI Analysis

●●SolidNiche GemCozy

Strömbom flocking model with bark mechanics in a Gymnasium wrapper.

Strengths
  • Implements the standard Strömbom et al. (2014) mathematical shepherding model accurately.
  • Exposed weights in flocking.py allow precise retuning for specific game builds.
  • Self-contained Python package with clean Gymnasium registration and vector/pixel obs modes.
Weaknesses
  • No pre-trained agents or benchmark scores provided to gauge difficulty baseline.
  • Narrow appeal limited to RL practitioners interested in multi-agent flocking problems.
Category
Target Audience

Reinforcement learning researchers and hobbyists

Similar To

PettingZoo · Multi-Agent Particle Environment · Google Research Football

Post Description

Created a new RL environment to test out world models. It's challenging enough for intermediate folks to experiment quickly.

Similar Projects

AI/MLMid

Gymnasium RL Environments in Julia

Julia port of Gymnasium environments, but Python's RL ecosystem already dominates this space.

Niche Gem
scascino4
202mo ago