Skip to content

AgileRL Documentation

AgileRL Documentation

Introduction

Getting Started
- AgileRL 2.0 Release Notes
Releases

Training

Evolutionary Hyperparameter Optimization
Off-Policy Training
On-Policy Training
Partially Observable Markov Decision Processes (POMDPs)
Offline Training
Multi-Agent Training
LLM Fine-Tuning
- Fused linear log-prob optimizations
- Saving and Loading LLM Checkpoints
Contextual Multi-Armed Bandits
Distributed Training
Evolvable Neural Networks
Creating Custom Algorithms
Debugging Reinforcement Learning

Tutorials

Gymnasium Single-agent Tutorials
PettingZoo Multi-agent Tutorials
Hierarchical Skills Tutorial
LLM Fine-Tuning Tutorials
Contextual Multi-arm Bandit Tutorials
- Iris with NeuralUCB
- PenDigits with NeuralTS
Creating Custom Networks Tutorials
- Building a Dueling Distributional Q Network
- Integrating Architecture Mutations Into SimBa

API

Algorithms
Components
Hyperparameter Optimization
- Mutation
- Tournament Selection
Modules
Networks
Training
Rollouts
- On-Policy Rollout Functions
Utils
Vector
- Petting Zoo Async Vector Environment
- Petting Zoo Vector Base Class
Wrappers

Development

GitHub
Discord
Contribute to AgileRL

Contextual Multi-arm Bandit Tutorials¶

These tutorials provide an introductory guide to using AgileRL contextual multi-armed bandit algorithms.

NeuralUCB - Iris

Pen digits image

NeuralTS - PenDigits

Iris with NeuralUCB

LLM Fine-Tuning with SFT and DPO

Copyright © 2023, AgileRL

Made with Sphinx and @pradyunsg's Furo