PyTorch Reinforcement Learning Integration

1 post · Dec 2025

What this trail covers

Together AI is integrating PyTorch Reinforcement Learning into its AI Native Cloud platform, providing a complete solution for building, training, and deploying agentic AI systems. This complements existing services like supervised fine-tuning, sandboxed environments, evaluation tools, model optimizations, and high-performance inference, enabling end-to-end development of advanced agentic systems. The platform now supports running TorchForge RL pipelines on Instant Clusters with distributed training and sandboxed environments.

All Together AI trails

Dec 2025How to run TorchForge reinforcement learning pipelines in the Together AI Native CloudThis post details the integration of TorchForge RL pipelines with Together AI's Instant Clusters, enabling distributed training and sandboxed environments. It highlights the use of InfiniBand/NVLink for low-latency GPU communication, preconfigured drivers and operators for consistent cluster bring-up, and heterogeneous scheduling for GPU and CPU-bound workloads. The post also showcases the integration of Together CodeSandbox and Together Code Interpreter as environments for tool-use and code execution within RL pipelines, providing a demo of a GRPO BlackJack training pipeline and a standalone integration for Code Interpreter.
Dec 2025Together AI and Meta partner to bring PyTorch Reinforcement Learning to the AI Native CloudThis post announces the integration of PyTorch Reinforcement Learning into the Together AI platform, enabling end-to-end development of agentic AI systems. It highlights the partnership with Meta's PyTorch team to build an open-source RL framework that complements existing services like supervised fine-tuning, sandboxed environments, evaluation tools, model optimizations, and high-performance inference. The integration aims to provide a complete platform for building, training, and deploying frontier agentic systems.
Advertisement