February 24 (Portal) – Meta Platforms Inc (META.O) announced on Friday that it is providing researchers with a new big language model, the core software of a new artificial intelligence system that’s fueling an AI arms race, while Big Tech companies are pushing to incorporate the technology into their products and impress investors.
The public battle for AI tech supremacy began late last year with the launch of Microsoft-backed OpenAI ChatGPT, prompting tech heavyweights from Alphabet Inc (GOOGL.O) to China’s Baidu Inc (9888.HK) to to present their own offers .
Metas LLaMA, short for Large Language Model Meta AI, will be available to researchers and entities associated with government, civil society and academia under a non-commercial license, a blog said.
Large language models sift through vast amounts of text to summarize information and generate content. For example, you can answer questions with sentences that read as if they were written by humans.
last update
Watch 2 more stories
The model, which Meta says requires “far less” processing power than previous offerings, is trained on 20 languages, with a focus on languages with Latin and Cyrillic alphabets.
“Today’s announcement by Meta appears to be a step in testing their Generative AI capabilities so they can implement them into their products in the future,” said Gil Luria, Senior Software Analyst at DA Davidson.
“Generative AI is a new application of AI that Meta has less experience with, but is clearly important to the future of their company.”
AI has emerged as a bright spot for investment in the tech industry, whose slowing growth has led to widespread layoffs and a slashing of experimental bets.
According to Meta, LLaMA could outperform competitors that examine more parameters or variables that the algorithm considers.
In particular, it said that a version of LLaMA with 13 billion parameters could outperform GPT-3, a more recent ancestor of the model ChatGPT is built on.
It described its 65 billion-parameter LLaMA model as “competitive” with Google’s Chinchilla70B and PaLM-540B, which are even larger than the model Google used to debut its Bard Chat-based search.
A meta-speaker attributed the performance to a larger amount of “clean” data and “architectural improvements” in the model that improved training stability.
Meta released the large language model OPT-175B in May last year, also aimed at researchers, and forming the basis of a new iteration of its chatbot, BlenderBot.
It later introduced a model called Galactica, which could write scientific articles and solve math problems, but quickly pulled the demo down after generating authoritative-sounding wrong answers.
Reporting by Yuvraj Malik and Eva Mathews in Bengaluru and Katie Paul in New York; Edited by Shailesh Kuber and Grant McCool
Our standards: The Trust Principles.