Take-aways:

  • Code Llama is an open-source LLM designed for code completion, generation, and testing.
  • Supported Languages: Python, C++, Java, and Bash.
  • Available for both research and commercial use at no cost.
  • Comes in 7B, 13B, and 34B parameter sizes for different performance needs.
  • Specialized Versions includes ‘Code Llama — Python’ and ‘Code Llama — Instruct’ for specific use-cases.
Meta just released its answer to GitHub Copilot, and it’s free
Meta’s specialized programming LLM can be run on a single GPU