Meta Open Sources Code Llama to Challenge AI Coding Assistants
Meta releases Code Llama, an open source AI model designed to generate, debug, and explain programming code across multiple languages. The new system builds upon the company's Llama 2 foundation to directly compete with existing tools like GitHub Copilot.
Meta releases Code Llama, a new open source artificial intelligence model that generates and explains computer code in English. This tool directly competes with established AI coding assistants like GitHub Copilot and Amazon CodeWhisperer by helping developers write and debug software across popular languages such as Python, C++, Java, and PHP.
The new system builds upon Meta's previously released Llama 2 text-generating model but focuses specifically on programming tasks. While Llama 2 possesses some basic ability to write code, Code Llama receives specialized training on 500 billion tokens of code-related data to ensure it produces much higher quality and more accurate programming results.
Meta offers Code Llama in several distinct variations to suit different developer needs, including a version specifically optimized for Python and another fine-tuned to understand natural language instructions. By making this technology openly available to the public, Meta aims to encourage community collaboration, accelerate innovation, and allow independent researchers to identify and fix potential security vulnerabilities.