Facebook's Mark Zuckerberg Unveils Meta's New Large Language Model

Mark Zuckerberg announced on Friday that Meta has trained and will release a new large language model to researchers.

‍

LLaMA, as the model is called, is intended to provide scientists and engineers with a framework for exploring AI applications such as answering questions and summarizing documents through the use of AI.

‍

It is important to note that Meta's release of its new model, which was developed by its Fundamental AI Research (FAIR) team, comes as large tech companies and well-capitalized startups alike race to announce the latest advances in artificial intelligence techniques, and integrate these into commercial products.

‍

The use of large language models underpins a number of applications, for example, OpenAI's ChatGPT, Microsoft's Bing AI, and Google's yet-to-be-released Bard.

‍

As Zuckerberg said in his post, LLM technology might eventually be able to solve math problems or be used to conduct scientific research.

‍

“We have seen a lot of promise with LLMs when it comes to generating text, facilitating conversations, summarizing written material, as well as solving math theorems and predicting protein structures,” Zuckerberg wrote in a blog post on Friday.

‍

There are several ways in which Meta's LLM distinguishes itself from competitive models, according to Meta.

‍

As a starting point, it mentions that it will come in several sizes, ranging from 7 billion parameters to 65 billion parameters. As the technology has advanced in recent years, larger models have been successful in expanding the capability of the technology, but they also cost more to operate, which is why researchers call this phase "inference."

‍

For example, OpenAI's Chat-GPT 3 has a parameter count of 175 billion.

‍

Meta has also stated that it will make its models available to the research community and that it is currently accepting applications from researchers to use its models. Google's LaMDA model and OpenAI's ChatGPT model are not transparent in terms of their underlying models.

‍

“Meta is committed to this open model of research, and we will make our new model available to the AI research community,” Zuckerberg wrote.

‍