Meta Releases LLaMA-AI Language Models for Scientific Research – IT Pro – News

Meta has released a set of AI language models to help researchers explore language models. LLaMA language modules are not yet publicly available. Meta has been working on AI language models for years, just like Google and OpenAI, for example.

The Large Meta AI Language Model, or LLaMA, consists of four AI language models of different sizes, all of which are relatively small. So they ask According to Meta Less computational power than other AI language models, which makes it more suitable for research purposes.

“Over the past year, language models have revealed new possibilities for creating creative texts, solving mathematical theories, predicting protein structures, and answering questions of reading comprehension,” Meta writes. By this, the company refers, among other things, to chat bots, for example, Google and Microsoft. However, Facebook’s parent company says it is difficult to conduct scientific research using these models because they are so large that it takes a lot of computing power to understand how and why they work.

Meta says this handicap makes optimizing robustness more difficult and “known issues such as bias, Poisoning and the potential spread of misinformation.” The company hopes that with smaller language models better research can be done on the matter. LLaMA language models are only made available after review to researchers involved in government agencies, civil society organizations, or scientific institutions. Last May, Meta released a language model Another for scientific research.