Submit a comment
Meta and Microsoft have partnered to develop Llama 2, a next-generation large language AI model aimed at both commercial and research purposes.
Meta logo at the World Economic Forum in Davos
On July 18, Meta - Facebook's parent company, announced the commercial version of the open-source artificial intelligence (AI) model Llama, competing with OpenAI and Google.
Meta and Microsoft have partnered to develop Llama 2, a next-generation large language AI model aimed at both commercial and research purposes.
Llama 2 will be distributed by Microsoft through the Azure cloud service running on the Windows operating system.
This model will provide information through files downloaded directly from Amazon Web Services, Hugging Face, and many other providers.
“Open source Llama 2 helps drive innovation in AI because this model allows many developers to build AI using new technology,” said Meta CEO Mark Zuckerberg.
Widespread adoption and free access for businesses to use a complex model like Llama will likely cause some AI chatbots like ChatGPT to lose their initial dominance.
The first version of Llama was already in direct competition with OpenAI's ChatGPT models and Google's Bard chatbot.
Llama 2 is trained on about 40% more data than its predecessor.
Sharing models developed by Meta researchers with different companies can spur innovation, uncover security vulnerabilities and reduce production costs, Zuckerberg said.
Not long ago, Meta also launched another AI model called I-JEPA that analyzes and directly completes the missing parts of an image, instead of just looking at a single pixel like many other general AI models.
In November 2022, Open AI's ChatGPT with the support of Microsoft launched the market, pushing the global technology world into the AI race.
According to Vietnam+
