Skip to content

MASA-Safe-RL

Overview

sacktock/MASA-Safe-RL

MASA-Safe-RL

sacktock/MASA-Safe-RL

Home
Get Started
Get Started
- Quick Start
- Core Concepts
  Core Concepts
- Basic Usage
Common API
Common API
- Constraints
  Constraints
  - Overview
  - Constrained Markov Decision Process (CMDP)
  - Linear Temporal Logic (LTL) Safety Constraint
  - Probabilistic Computation Tree Logic (PCTL) Constraint
  - Step-wise Probabilistic Constraint
  - Reach-avoid Constraint
  - Multi-Agent Constraints
    Multi-Agent Constraints
    
    Overview
    
    Constrained Markov Game (CMG)
    
    Alternating-Time Temporal Logic (ATL) Safety
- Wrappers
  Wrappers
- Metrics
  Metrics
  - Overview
  - Logging
- Linear Temporal Logic (LTL)
  Linear Temporal Logic (LTL)
- Probabilistic Computation Tree Logic (PCTL)
  Probabilistic Computation Tree Logic (PCTL)
Environments
Environments
- Multi Agent
  Multi Agent
  - Overview
  - Gridworlds
    Gridworlds
    
    Overview
    
    Capture the Flag
    
    Clean Up
    
    Markov Stag Hunt
  - Matrix Games
    Matrix Games
    
    Overview
    
    Bertrand
    
    Chicken
    
    Congestion
    
    Dynamic Public Goods Game
    
    Inspection
- Single Agent
  Single Agent
  - Overview
  - Cartpole
  - Safety Gridworlds
    Safety Gridworlds
    
    Overview
    
    Island Navigation
    
    Conveyor Belt
    
    Sokoban
  - Pacman
  - Gridworlds
    Gridworlds
    
    Overview
    
    Bridge Crossing
    
    Colour Grid World
    
    Colour Bomb Grid World
  - Media Streaming
Algorithms
Algorithms
- Algorithms Overview
- Tabular Algorithms
  Tabular Algorithms
  - Overview
  - Q Learning
  - Q Learning Lambda
  - LCRL
  - SEM
  - RECREG
- Multi-agent Tabular Algorithms
  Multi-agent Tabular Algorithms
- On-Policy Algorithms
  On-Policy Algorithms
Shielding
Shielding
- Probabilistic Shielding
- Winning-region safety-game shielding
- Shielded Algorithms
  Shielded Algorithms
Tutorials
Tutorials
- Basics
  Basics
- Constraints
  Constraints
  - Overview
  - Constraints Tour
  - LTL-Safety
    LTL-Safety
    
    Overview
    
    LTL Safety Colour Bomb
- Wrappers
  Wrappers
  - Overview
  - Vectorization and Normalization
- Environments
  Environments
  - Overview
  - Create a New Environment
- Baselines
  Baselines
- Shielding
  Shielding
- Multi-Agent
  Multi-Agent
  - Overview
  - Multi-Agent CMG

On-Policy Algorithms¶

This section covers MASA's on-policy actor-critic methods. The currently implemented and registered algorithms in this part of the codebase are A2C and PPO.