Meta Unveils Updated AI Version
In a partnership move, tech giants Meta and Microsoft have joined forces to roll out Llama 2, a large-scale, open-source language model from Meta's suite of tools. This model is set to be featured on Microsoft's Windows operating system and its cloud service, Azure.
The collaboration was made official on July 18, clarifying that Llama 2 is available free of charge for commercial and research purposes, and is specifically optimized for Windows.
This confirmation substantiated prior speculations that Llama 2 was engineered to enable businesses and researchers to build applications on Meta’s AI technological stack.
According to Meta, Llama 2 was trained using 40% more data from publicly available online sources and has twice the processing capacity of its predecessor, Llama 1.
Despite surpassing many competitor open-source large language models in coding capability, proficiency, reasoning, and performance on knowledge tests, Meta admitted that Llama 2’s efficiency does not measure up to its closed-source rivals, such as OpenAI's GPT-4, based on findings from its research paper.
In an Instagram post on July 18, Mark Zuckerberg, CEO of Meta, stated that Llama 2 "provides businesses and researchers the ability to build upon our next generation large language model."
In response to the limited release of Llama 1 in February, Meta reported being "taken aback" by the demand, with over 100,000 requests for access. This model was later leaked on the 4chan imageboard website.
However, the popularity of Llama 1 was outshone by ChatGPT, which reportedly attracted upwards of 100 million users in the first three months after its launch, according to a Reuters report from February.
Through this partnership, Microsoft now backs two key players in the AI field. As of 2023, it had invested a total of $13 million in OpenAI, according to a report by Fortune published in January.
This confirmation substantiated prior speculations that Llama 2 was engineered to enable businesses and researchers to build applications on Meta’s AI technological stack.
According to Meta, Llama 2 was trained using 40% more data from publicly available online sources and has twice the processing capacity of its predecessor, Llama 1.
Despite surpassing many competitor open-source large language models in coding capability, proficiency, reasoning, and performance on knowledge tests, Meta admitted that Llama 2’s efficiency does not measure up to its closed-source rivals, such as OpenAI's GPT-4, based on findings from its research paper.
In an Instagram post on July 18, Mark Zuckerberg, CEO of Meta, stated that Llama 2 "provides businesses and researchers the ability to build upon our next generation large language model."
In response to the limited release of Llama 1 in February, Meta reported being "taken aback" by the demand, with over 100,000 requests for access. This model was later leaked on the 4chan imageboard website.
However, the popularity of Llama 1 was outshone by ChatGPT, which reportedly attracted upwards of 100 million users in the first three months after its launch, according to a Reuters report from February.
Through this partnership, Microsoft now backs two key players in the AI field. As of 2023, it had invested a total of $13 million in OpenAI, according to a report by Fortune published in January.
