areal

We have hosted the application areal in order to run this application in our online workstations with Wine or directly.

Quick description about areal:

AReaL is an open source, fully asynchronous reinforcement learning training system. AReal is designed for large reasoning and agentic models. It works with models that perform reasoning over multiple steps, agents interacting with environments. It is developed by the AReaL Team at Ant Group (inclusionAI) and builds upon the ReaLHF project. Release of training details, datasets, and models for reproducibility. It is intended to facilitate reproducible RL training on reasoning / agentic tasks, supporting scaling from single nodes to large GPU clusters. It can streamline the development of AI agents and reasoning systems. Support for algorithm and system co-design optimizations (to improve efficiency and stability).

Features:

• Fully asynchronous RL architecture (rollouts decoupled from training)
• Ability to scale from one node up to 1,000+ GPUs
• Flexible customization for multi-turn agentic rollout workflows
• Integration with agentic tool frameworks / pipelines
• Support for algorithm and system co-design optimizations (to improve efficiency and stability)
• Release of training details, datasets, and models for reproducibility

Programming Language: Python.
Categories: