Feb 24 (Reuters) – Meta Platforms Inc ( METAO ) said on Friday it is launching a new AI-based big language model for the research community, becoming the latest company to join the artificial intelligence race.
The battle for control of artificial intelligence technology, until recently on the back burner, began late last year with the launch of Microsoft-backed ChatGPT software, which has attracted tech giants from Alphabet Inc ( GOOGL.LO ) to China’s Baidu. (9888. HK) to create your own show.
MetaLama, which stands for Large Language Model MetaAI, will be available under a non-commercial license to researchers, government agencies, civil society and academia, he said in a blog post.
The company will provide built-in code for users to modify the form and use it in search-related use cases. The model, which Matt says requires “very little” computing power, was trained in 20 languages, with an emphasis on the Latin and Cyrillic alphabets.
Jill Lurie, Senior Software Analyst, D.A. Davidson.
“Generative AI is a new application of AI that Meta has little experience with, but is clearly important to the future of their business.”
Artificial intelligence has become a bright spot for investment in a tech industry whose slow growth has been fueled by mass layoffs and reduced rates of experimentation. Meanwhile, Microsoft Corp. ( MSFT.O ), Baidu and Alphabet’s Google are integrating their advanced AI language engines into more collective products such as search.
Last May, Meta released a larger language model, the OPT-175B, also aimed at researchers, which formed the basis for a new iteration of the Blenderbot chatbot.
A Galactica model was later launched, which it said could write scientific papers and solve math problems, but its demo was later pulled because it repeated official material.