Meta Open-Sourced Llama 2, Its Newest Generative AI Model

IBL News | New York

Meta, in partnership with Microsoft, this week announced it is open-sourcing the Llama 2 generative AI model and making it available free of charge for research and commercial use.

The company is including model weights and starting code for the pre-trained model and conversational fine-tuned versions too.

Starting this week, Llama 2 is available in the Azure AI model catalog, enabling developers using Microsoft Azure to build with it. It is also available through Amazon Web Services (AWS), Hugging Face, and other providers.

Llama 2 is optimized to run locally on Windows.

Among other joined initiatives, Meta and Microsoft already collaborated to scale the adoption of PyTorch —which is now the leading framework for AI — on Azure.

The Llama 2 model has been fine-tuned and red-teamed — tested for safety — through internal and external efforts.

The team, which worked to generate adversarial prompts to facilitate model fine-tuning, plans to release updated fine-tuned models.

Meta created this guide as a resource to support developers with best practices for development and safety evaluations.

In addition, the company launched a new partnership program for academic researchers that aims to deepen its understanding of the responsible development and sharing of large language models. Researchers may apply to join a community of practitioners to share learnings on this important topic, and the community will form a research agenda to pursue going forward.