Red Hat Reveals Major Enhancements to Red Hat Enterprise Linux AI

Red Hat has announced the release of Red Hat Enterprise Linux AI (RHEL AI) 1.2, just weeks after the initial version 1.0. This update introduces significant enhancements aimed at improving the development, testing, and deployment of large language models (LLMs). RHEL AI is designed to facilitate generative AI model processes while making LLM training more cost-effective. The new version integrates IBM Research's Granite LLM family and the InstructLab alignment tools, promoting collaborative model development through the open-source InstructLab AI project. Key features include Retrieval-Augmented Generation (RAG) for better access to external knowledge, expanded hardware support for Lenovo servers, and a technology preview for AMD Instinct Accelerators. RHEL AI 1.2 also supports deployment on major cloud platforms like Azure and Google Cloud, alongside existing support for AWS and IBM Cloud. New functionalities such as "Periodic Checkpointing" allow users to save training progress, while the PyTorch Fully Sharded Data Parallel (FSDP) feature significantly reduces training times. Red Hat aims to make LLM training accessible to a broader range of users, not just data scientists. With this rapid release cycle, Red Hat is reinforcing its commitment to the enterprise AI market, urging users to upgrade from the deprecated version 1.1 within 30 days.

- Red Hat released RHEL AI 1.2, enhancing LLM development and deployment.

- The update includes new features like Retrieval-Augmented Generation and Periodic Checkpointing.

- RHEL AI 1.2 supports additional hardware and major cloud platforms.

- The software aims to make AI model training accessible to non-data scientists.

- Users are encouraged to upgrade from the deprecated version 1.1 within 30 days.