Meta has just launched Llama 2, which they claim to be the ‘next generation of our open source large language model (LLM).’ This release puts Llama 2 in competition with ChatGPT (OpenAI) and Bard (Google) as a potential rival. Meta has made Llama 2 accessible for research and commercial purposes, offering it free-of-cost.
Llama 2, as described by Hugging Face, one of Meta’s global partners and supporters, is a family of state-of-the-art open-access language models released by the tech giant.
Key features of Llama 2 include:
- It was trained on 40% more data than its predecessor, Llama 1, and boasts double the context length. The model comes in three sizes: 7B, 13B, and 70B.
- Meta claims that Llama 2 outperforms other open source LLMs in various external benchmarks, including reasoning, coding, proficiency, and knowledge tests.
- The Llama-2-chat, a ‘fine-tuned’ model, is pre-trained on publicly available online data sources and utilizes publicly available instruction datasets, along with over 1 million human annotations.
Numerous companies from around the world, including Microsoft, Amazon Web Services (AWS), IBM, Spotify, Shopify, Jio, Qualcomm, Nvidia, Accenture, Zoom, T-Mobile, and LG, among others, support the Llama 2 project.
Llama 2 is readily available through respective providers and can be found in the Azure AI model catalog, with optimized support for local Windows running.
Also Read: Jury Awards $18.8 Million in Cancer Lawsuit Against Johnson & Johnson’s Baby Powder
.' This release){kind=link}