Mita launches the "Code Llama" artificial intelligence assistant for developers.. What's new?
- Meta announces "Code Llama," a language template designed specifically for developers that enhances code generation and streamlines workflow.
- Code Llama is based on the "Llama 2" model and three versions have been released to meet different needs of developers.
- Although Code Llama does not currently support certain smart languages such as Solidity, there are initiatives focused on this aspect, showing interest in the integration of AI and programming.
Meta, Facebook's parent company, unveiled "Code Llama" on August 24, a large language model (LLM) designed specifically for developers. This smart tool aims to enhance code generation, streamline workflows, and provide an accessible entry point for budding programmers.
Building on the open source language model "Llama 2," first launched in July, Meta has customized its capabilities to meet the needs of the programming community. Code Llama is not only a code generator, it is a comprehensive tool capable of discussing code and facilitating debugging. It is characterized by its ability to generate code and related natural language, as well as ingest both types of information.
Three versions of "Code Llama" were released, each designed to meet specific needs: The smallest version, 7B, could run on a single GPU. While the larger 34B model promises better results and improved markup directives. Dedicated versions of Python have also been designed and others focused on understanding human instructions effectively.
This development comes in light of the growing interest in combining artificial intelligence and programming, especially in the field of blockchain. Although Code Llama does not currently support popular smart contract languages such as Solidity, other initiatives, such as ChainGPT and Decrypted, are working on similar features.
The launch of Code Llama marks a major step towards the integration of AI and software development. And as the age of technology advances, tools like this stand out as evidence of the importance of AI in shaping the future of programming and software design.