Meta says Llama’s usage grew tremendously due to the power of open source
Meta says Llama’s usage grew tremendously due to the power of open source
Llama is ‘approaching 350 million downloads to date.’
Meta has published an update on how its Llama large language models are performing, and they’re apparently doing so well that they’re now “approaching 350 million downloads to date.” That’s 10 times more than the downloads it accumulated compared to the same time last year. Approximately 20 million of those downloads took place in the last month alone, after the company released Llama 3.1, its latest LLM that it says can now rival OpenAI’s and Anthropic’s.
The monthly usage of Llama grew ten times from January to July this year for some of Meta’s largest cloud service providers, the company said. From May to July, in particular, hosted Llama usage on its cloud partners more than doubled by token volume. In addition to Amazon Web Services (AWS) and Microsoft Azure, the company teamed up with Databricks, Dell, Google Cloud, Groq, NVIDIA, IBM watsonx, Scale AI and Snowflake, among others, to make its LLMs more readily available to developers.
Meta attributed Llama’s growing usage to “the power of open source.” It said that making the LLMs available to all allowed it to see “a vibrant and diverse AI ecosystem come to life where developers have more choice and capability.” When Meta released Llama 3.1, Mark Zuckerberg extolled the virtues of open source AI, calling it the “path forward.” He also talked about the steps Meta is taking to make open source AI the industry standard.
In its latest report, it gave examples of how companies are using Llama for their needs. AT&T apparently uses Llama to fine-tune customer searches, while DoorDash relies on the LLM to make tasks easier for its software engineers. Llama also generates the real-time reactions of the creatures called “Dots” in Niantic’s Tamagotchi-like game Peridot. Meanwhile, Zoom uses Llama (along with other LLMs) to power its AI Companion that can summarize meetings and make smart recordings. “The innovation has been broad and rapid, from start-ups pushing new boundaries to enterprises of all sizes using Llama to build on-premises or through a cloud service provider,” Meta wrote. “Industry is building and innovating with Llama, and we’re even more excited for what’s to come.”
(2)
