Meta and Microsoft open source LLaMA 2 to challenge GPT-4

In a strategic move to rival OpenAI's ChatGPT, Meta has announced the open-sourcing of its large language model, LLaMA 2, in partnership with Microsoft.

The decision, unveiled during Microsoft's Inspire event, aims to provide businesses, startups, and researchers with access to powerful AI tools for commercial and research purposes, sharpening competition in the language model space.

By making LLaMA 2 freely available, Meta aims to encourage experimentation and collaboration in the AI community. The model, trained on a much larger dataset than its predecessor LLaMA 1, incorporates information from publicly available online sources.

Meta claims that LLaMA 2 outperforms other large language models, including Falcon and MPT, in areas such as reasoning, coding, qualification and knowledge tests.

LLaMA 2 will be available through Microsoft's Azure platform, as well as through AWS, Hugging Face and other providers. Meta's partnership with Qualcomm will expand the scope of LLaMA by extending the technology to laptops, phones, and headsets, which will make it possible to create AI-based applications that work independently of cloud services starting in 2024.