Microsoft made the announcement on Tuesday at its annual Inspire event that it has signed a contract with Mark Zuckerberg’s Meta to integrate the artificial intelligence (AI) model known as LLama 2 into the cloud-computing platform known as Microsoft Azure.
Meta has been doing phenomenal work innovating in the open models, and Llama has captured the imagination of what open-source and AI foundation models can do. We are very excited today is the announcement of Meta’s Llama 2 coming to Azure and Windows.
Microsoft CEO Satya Nadella said during the Azure AI presentation
According to the press release published by Meta, the choice to open up LLaMA was made as a means of providing enterprises, startups, and academics with access to additional AI tools, therefore enabling the community to experiment together. According to Meta, LLaMa 2 was trained on 40 percent more data compared to LLaMa 1, which includes information from “publicly available online data sources”. When it comes to reasoning, coding, competency, and knowledge assessments, it claims that it “outperforms” rival LLMs like Falcon and MPT.
The public release of Llama was announced by Mark Zuckerberg, CEO of Meta, in the month of February. Llama is designed to assist scientists and engineers in exploring applications for artificial intelligence, such as answering queries and summarizing materials. Even while Meta will not directly gain any money from the partnership with Microsoft, the company stands to possibly profit by increasing the number of customers and developers who have access to its proprietary artificial intelligence software. Hugging Face is a well-known service that many AI researchers make use of, and Amazon Web Services will also make Llama 2 accessible to its customers.
Because of this agreement, it is clear that Microsoft is supporting an AI language model that was not created by OpenAI, which is a big thing for the software giant. The parent corporation of Facebook, Instagram, and Threads, Meta, has shifted its focus this year from the metaverse and NFT initiatives to artificial intelligence. Meta just released CM3leon, a new text-to-image model that can create graphics based on word inputs, last week.
We want to make it easier for people to access information and use devices in their preferred language, and today we’re announcing a series of artificial intelligence models that could help them do just that.
Meta said at the time
