Library/AReaLForked

inclusionAI/AReaL

AReaL

Lightning-Fast RL for LLM Reasoning and Agents. Made Simple & Flexible.

View on GitHub↗Upstream inclusionAI/AReaL↗

Builder

inclusionAI

inclusionAI • individual

Stars

4,967

Using upstream star count

Forks

446

Using upstream fork count

Open Issues

Activity Score

0/100

0 commits in 30d

Created

Feb 24, 2025

Project creation date

README Summary

AReaL is a lightning-fast reinforcement learning framework specifically designed for training large language models on reasoning tasks and agent behaviors. It provides a simple and flexible interface for implementing RL algorithms to improve LLM performance on complex reasoning problems.

AI Dev Skills

Unmapped

Reinforcement Learning for LLMsLarge Language Model TrainingLLM Reasoning SystemsAI Agent DevelopmentPolicy Gradient MethodsReward Model TrainingLanguage Model Fine-tuning

Taxonomy

AI Trends

Agentic AI LLM Reasoning Reinforcement Learning from Human Feedback AI Agents

Deployment Context

Self-hosted Cloud API

Modalities

Text

Skill Areas

Reinforcement Learning for LLMs Large Language Model Training LLM Reasoning Systems AI Agent Development Policy Gradient Methods Reward Model Training Language Model Fine-tuning

Use Cases

LLM Reasoning Enhancement AI Agent Training Reinforcement Learning from Human Feedback Multi-step Reasoning Tasks Decision-making Agent Development

Recent Activity

Updated 28 days ago

7 Days

30 Days

90 Days

Quality

research

Quality: medium
Maturity: research

PM Skills

Developer Platform

Languages

Python100.0%

Timeline

Project created: Feb 24, 2025
Forked: Mar 13, 2026
Your last push: 28 days ago
Upstream last push: 7 days ago
Tracked since: Mar 17, 2026

Similar Repos

pgvector cosine similarity · $0

Loading…