Meta reveals its largest models of artificial intelligence
On Tuesday, Meta Platforms released the largest version of its AI model, Llama 3, which is mostly free, multilingual, and has public performance metrics that rival paid models from rival companies such as OpenAI.Meta, the owner of Facebook, said in a blog post and in a research paper announcing the release that the new (Llama 3) model can speak eight languages, write high-quality computer codes and solve more complex mathematical problems than previous versions.
Last year's version is dwarfed by this one, which includes 405 billion metrics or variables that algorithms take into account to generate responses to user queries, although it is still smaller than competitors' leading models. There is news that OpenAI's GPT-4 model contains one trillion variables, and Amazon is investing in a model that contains two trillion.
Hectic race
The release comes as technology companies race to show that their growing portfolios of large, resource-hungry language models can yield big enough gains in problematic areas like advanced reasoning to justify the huge sums invested in them.
Meta Platforms stated that in addition to its flagship model, Llama 3, it will launch two updated versions of its short models, one of which includes eight billion variables, and the second includes 70 billion, which it had introduced in the spring.
Meta largely allows developers to use Llama models for free, a strategy that CEO Mark Zuckerberg says will pay off in the form of innovative products and greater engagement on the company's core social media networks. However, the costs involved have raised some eyebrows among investors.
0 Comments