Meta has made its new massive language model, Code Llama (CL), ready for download for researchers and developers. It is a generative AI & language model that is free to use and features 70 billion parameters. It was originally announced in August 2023.
Today we’re releasing Code Llama 70B: a new, more performant version of our LLM for code generation — available under the same license as previous Code Llama models.
— AI at Meta (@AIatMeta) January 29, 2024
“We’re open-sourcing a new and improved Code Llama, including a larger 70B parameter model. Writing and editing code has emerged as one of the most important uses of AI models today,” said Meta CEO Mark Zuckerberg in a post on Facebook.
What makes Code Llama so powerful?
According to Meta, it is the most powerful and effective model in the Llama line. It is available in three variants: CodeLlama – 70B, the foundational code model, CodeLlama – 70B – Python, 70B specialized for Python, and Code Llama – 70B – Instruct 70B. The last variant is an improved version that combines self-instruction code synthesis with human instruction.
Code Llama is built on Llama and is capable of producing code, and of writing Python, PHP, C++, and Java code. It has scored 67.8 on the generative AI benchmark, HumanEval. It is said to be optimised for code generation by Meta, and its best feature is that it is an open-source model.