Mark Zuckerberg says that Meta’s Llama models have hit 1B downloads

MT HANNACH
3 Min Read
Disclosure: This website may contain affiliate links, which means I may earn a commission if you click on the link and make a purchase. I only recommend products or services that I personally use and believe will add value to my readers. Your support is appreciated!

In a brief message Tuesday morning on the threads, the meta-PDG Mark Zuckerberg said that the family, Llama, “open” Model Family, Llama of the company, hit 1 billion downloads. This represents 650 million downloads at the beginning of December 2024 – an increase of 153% over a period of about three months.

Lama, who feeds the AI ​​assistant from Meta, Meta AiThrough the different platforms of the technology giant, including Facebook, Instagram and WhatsApp, is part of Meta’s candidacy, to promote a large-scale AI product ecosystem. The company manufactures models, as well as the tools necessary to refine and personalize them, available free of charge under a owner license.

Some developers and companies have challenged the Lama license conditionswhich are somewhat commercially restrictive. However, Llama was a great success since its launch in 2023 despite this. Companies such as Spotify, AT&T and Doordash use LLAMA models in production today.

This does not want to suggest that Meta did not face backhands.

Lama is at the center of a AI copyright trials Who accuses Meta of forming a certain number of models on electronic books protected by copyright without authorization. Several EU countries Forced Meta to postpone – and in some cases completely cancel – his lama launch plans concerning data confidentiality problems. And Llama’s performance was overwhelmed by models such as the Chinese Deepseek laboratory R1.

Meta would have scrambled to set up “war rooms” to apply Deepseek learning to the development of Llama, and the company recently declared that it spend up to $ 80 billion On AI projects this year.

Meta plans to launch Several models of Lama in the coming months, including “reasoning” models in the direction of the Openai O3-min and models with multimodal capabilities natively. Zuckerberg also “Agent” characteristicssuggesting that some of these models will be able to take measures independently.

“I think it could very well be the year when Llama and Open Source become the most advanced and used AI models,” Zuckerberg said when Meta’s quarter 2024 results in January. “”[O]your goal for [Llama this year] is to direct. “”

We are sure to know more to LlamaconThe first generative conference of Meta AI developers, which should take place on April 29.

Share This Article
Leave a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

What do you like about this page?

0 / 400