Meta today launched Code Llama, an AI tool built on its open-source large language model (LLM) Lllama 2, made for coders and developers.
According to Meta's blog post, Code Llama is designed to speed up workflows and make coding easier for beginners.
"Code Llama has the potential to be used as a productivity and educational tool to help programmers write more robust, well-documented software," the company explained.
The new AI tool can generate code and natural language about code from both natural language and code prompts.
It also supports a wide variety of programming languages, such as Python, Java, C++, Typescript (Javascript), PHP, C#, Bash, and more.
Code Llama will be released in 7B, 13B, and 34B parameters, with each model trained with 500B tokens of code and data and serving their own specific purpose.
The 7B model can be served on a single GPU and together with 13B, are faster for tasks that don't require high latency. Meanwhile, the 34B model gives back the best results and offers superior coding assistance.
"Programmers are already using LLMs to assist in a variety of tasks. The goal is to make developer workflows more efficient so that they can focus on the most human-centric aspects of their job, rather than repetitive tasks," the company concluded.
Edited by Nikola Djuric
-author-avatar_thumbnail.jpg){kind=avatar}