🏆 Code Llama 70B from Meta: the Best Open-Source Model Surpassing GPT-4
Meta recently released a new artificial intelligence model called Code Llama 70B, which surpasses its predecessor, GPT-4. In the Human Eval test, Code Llama 70B scored 67.8 points, while GPT-4 scored 67.0 points. This makes it the most powerful model for code generation available in the open.
Code Llama 70B is available in three versions, each free for both research and commercial use:
- Base code model.
- Python, 70B, specialized for the Python language.
- Instruct 70B, adapted for understanding instructions in natural language.
🔗 Blog post ai.meta.com/blog/code-llama-large-language-model-coding
📝 Paper: ai.meta.com/research/publications/code-llama-open-foundation-models-for-code
👨💻 Github: github.com/facebookresearch/codellama
