DeepSeek Coder

DeepSeek Coder is a family of open-source large language models specifically designed for code generation and related tasks, aiming to provide a powerful and accessible alternative to closed-source models. Research focuses on improving model performance through techniques like Mixture-of-Experts architectures and repeated sampling during inference to boost problem-solving capabilities, even surpassing some closed-source competitors on certain benchmarks. The availability of these open-source models significantly advances code intelligence research and facilitates broader access to advanced code generation tools for both research and commercial applications.

Papers