Meta introduced the next generation open source large language model (LLM), Llama 2.
This is in line with the expansion of Meta's partnership with Microsoft, which is the "preferred partner" for Llama 2.
According to Meta, Llama is available free for research and commercial use.
"We're opening access to Llama 2 with the support of a broad set of companies and people across tech, academia, and policy who also believe in an open innovation approach to today's AI technologies," the company wrote on a blog post.
What's Inside Meta's Llama 2?
Llama 2 model was trained on 40% more data than its predecessor, Llama. Additionally, the new model has doubled the context length.
As for benchmarks, Llama 2 outperforms other open source language models in the scope of reasoning, coding, proficiency, and knowledge tests.
Most importantly, Llama 2 reinforces learning from human feedback which helps Meta and Microsoft ensure its safety and efficiency.
Meta's Llama 2 Availability
Meta also announced that Llama 2 will be available through Microsoft's Azure platform, AWS, Hugging Face, and other providers.
"It is also optimized to run locally on Windows, giving developers a seamless workflow as they bring generative AI experiences to customers across different platforms," Meta explained.
Meta also shared that its partnership with Microsoft extends outside AI, including the metaverse to "deliver immersive experiences" in the future of work and play.
As of writing, Meta is encouraging researchers to join a community of practitioners to share learning on AI, with the community forming a research agenda to pursue the same goal moving forward.
Related Article : Microsoft Launches New Program to Certify Users on Generative AI
