Facebook’s parent company Meta has released the latest iteration of its AI model called Llama 3.1. What makes this announcement significant is that this new model is “by far the most advanced open source model” available, and is the largest one developed by Meta to date. The launch of Llama 3.1 comes as Meta looks to position itself as a major player alongside tech giants like Google, Amazon, and startups such as OpenAI and Anthropic.
Company CEO Mark Zuckerberg made the announcement in a video on Instagram.
“We have got another big AI release today,” Zuckerberg said in the video. He said that “Meta is on track to becoming the most used AI assistant in the world by the end of the year”, adding that there are hundreds of millions of people using it everyday. The tech CEO also said that the AI model will be available in more countries soon.
How Llama 3.1 is different from Llama 3
Zuckerberg added that Llama 3.1 will support more languages, offer more reasoning and is “just better overall” as compared to the predecessor. He noted that the company is adding a feature that can generate images in which ‘users can put themselves in’.
As per the company, the new model is being released with three training parameters: Llama 3.1 with 405 billion parameters, the first frontier-level open source AI model, as well as improved Llama 3.1 70 billion and 8 billion models.
“In addition to having significantly better cost/performance relative to closed models, the fact that the 405B model is open will make it the best choice for fine-tuning and distilling smaller models,” the company said. Zucekrberg also noted that Meta will also release a tool that will allow customers their own AI models.
Meta is working with Amazon and Nvidia for developer access
Similar to Meta’s release of Llama 2, the company is partnering with other tech companies that will offer their customers access to Llama 3.1 via their respective cloud computing platforms.
“Beyond releasing these models, we’re working with a range of companies to grow the broader ecosystem. Amazon, Databricks, and Nvidia are launching full suites of services to support developers fine-tuning and distilling their own models,” Meta added.
Llama 3.1’s largest version (with 405 billion parameters) has been trained with over 16,000 of Nvidia’s H100 GPUs.
Company CEO Mark Zuckerberg made the announcement in a video on Instagram.
“We have got another big AI release today,” Zuckerberg said in the video. He said that “Meta is on track to becoming the most used AI assistant in the world by the end of the year”, adding that there are hundreds of millions of people using it everyday. The tech CEO also said that the AI model will be available in more countries soon.
How Llama 3.1 is different from Llama 3
Zuckerberg added that Llama 3.1 will support more languages, offer more reasoning and is “just better overall” as compared to the predecessor. He noted that the company is adding a feature that can generate images in which ‘users can put themselves in’.
As per the company, the new model is being released with three training parameters: Llama 3.1 with 405 billion parameters, the first frontier-level open source AI model, as well as improved Llama 3.1 70 billion and 8 billion models.
“In addition to having significantly better cost/performance relative to closed models, the fact that the 405B model is open will make it the best choice for fine-tuning and distilling smaller models,” the company said. Zucekrberg also noted that Meta will also release a tool that will allow customers their own AI models.
Meta is working with Amazon and Nvidia for developer access
Similar to Meta’s release of Llama 2, the company is partnering with other tech companies that will offer their customers access to Llama 3.1 via their respective cloud computing platforms.
“Beyond releasing these models, we’re working with a range of companies to grow the broader ecosystem. Amazon, Databricks, and Nvidia are launching full suites of services to support developers fine-tuning and distilling their own models,” Meta added.
Llama 3.1’s largest version (with 405 billion parameters) has been trained with over 16,000 of Nvidia’s H100 GPUs.
Facebook-parent Meta releases “the most advanced open source” AI model yet #Facebookparent #Meta #releases #advanced #open #source #model
Source Link: https://timesofindia.indiatimes.com/technology/tech-news/facebook-parent-meta-releases-the-most-advanced-open-source-ai-model-yet/amp_articleshow/111969392.cms
Facebook-parent Meta releases “the most advanced open source” AI model yet
Facebook’s parent company Meta has relea… – BLOGGER – WP1, Advanced, Facebookparent, Meta, model, open, Releases, Source
