Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Meta unveils a new, more efficient Llama model | TechCrunch

Dec 06, 2024 - techcrunch.com
Meta has introduced a new addition to its Llama family of generative AI models, Llama 3.3 70B. According to Ahmad Al-Dahle, VP of generative AI at Meta, this text-only model delivers the same performance as Meta's largest Llama model but at a lower cost. It outperforms Google’s Gemini 1.5 Pro, OpenAI’s GPT-4o, and Amazon’s Nova Pro on several industry benchmarks. The model is available for download from the AI dev platform Hugging Face and other sources, including the official Llama website.

However, Meta's terms limit how certain developers can use Llama models; platforms with more than 700 million monthly users must request a special license. Despite this, Llama models have been downloaded over 650 million times. Meta is also investing in its infrastructure to train and serve future generations of Llama models, with plans to build a $10 billion AI data center in Louisiana.

Key takeaways:

  • Meta has introduced a new AI model, Llama 3.3 70B, which delivers the performance of its largest model, Llama 3.1 405B, at a lower cost.
  • The new model outperforms Google’s Gemini 1.5 Pro, OpenAI’s GPT-4o, and Amazon’s Nova Pro on several industry benchmarks.
  • Despite concerns about the 'open' nature of Llama models, they have been downloaded over 650 million times and power Meta's AI assistant, which has nearly 600 million monthly active users.
  • Meta is investing heavily in its compute infrastructure, with plans to build a $10 billion AI data center in Louisiana to train future generations of Llama models.
View Full Article

Comments (0)

Be the first to comment!