Meta has announced the release of LLaMA 2, an upgrade to the company’s open-source large language model (LLM). The project marks another milestone in Meta’s partnership with Microsoft in developing AI tools.
LLaMA 2 (Large Language Model Meta AI) is a collection of pretrained LLMs ranging in scale from 7 billion to 70 billion parameters.
The models are trained on 2 trillion tokens and have a context window of 4,096 tokens, and over 1 million human annotations. Tokens in AI model training are the basic units of text or code that an LLM uses to process and generate language. Tokens can be characters, words, subwords, or other segments of text or code.
According to Meta’s blog post, the company trained the model on 40% more data than its predecessor, and the AI model “outperforms” other LLMs like Falcon and MPT when it comes to reasoning, coding, proficiency, and knowledge tests.
The upgrade comes after the initial release of the AI model in February.
Open for research and commercial use
With the introduction of a new generation of its AI model, the company is opening it up for research and commercial use free of charge.
“We’re opening access to Llama 2 with the support of a broad set of companies and people across tech, academia, and policy who also believe in an open innovation approach to today’s AI technologies,” said the company.
“Opening access to today’s AI models means a generation of developers and researchers can stress test them, identifying and solving problems fast, as a community. By seeing how these tools are used by others, our own teams can learn from them, improve those tools, and fix vulnerabilities,” states Meta.
Meta is standing firmly behind the partnership with Microsoft in developing AI tools, saying that “people and businesses have benefited from the longstanding partnership” between the companies.
Your email address will not be published. Required fields are markedmarked