Yi-Coder

Yi-Coder is a family of open-source large-scale language models (LLMs) that provide state-of-the-art coding performance with less than 10 billion parameters. It comes in two sizes—1.5B and 9B parameters—in base and chat versions and is designed for efficient inference and flexible training. Yi-Coder -9B was trained on an additional 2.4 trillion high-quality tokens on GitHub's code base-level code corpus and code-related data filtered from CommonCrawl. Yi-Coder excels at a variety of programming tasks, including basic and competitive programming, code editing and warehouse-level completion, long-context understanding, and mathematical reasoning.

Demand group:

" Yi-Coder is suitable for software developers, programming competition participants and AI researchers. It can improve the efficiency of code writing, help solve complex programming problems, and provide a research basis for the application of AI in the field of programming."

Example of usage scenario:

Developers use Yi-Coder to automatically complete codes and improve development efficiency

Programming competition contestants use Yi-Coder to solve algorithm problems and gain higher rankings

AI researchers use Yi-Coder to conduct research on code generation and reasoning capabilities

Product features:

Supports high-quality token pre-training in 52 major programming languages

Long context modeling: the maximum context window is 128K tokens, enabling project-level code understanding and generation

Excellent performance in models with less than 1 billion parameters, comparable to larger models

On the LiveCodeBench platform, Yi-Coder -9B-Chat achieved a pass rate of 23.4%, surpassing other models

In CodeEditorBench, Yi-Coder -9B performed well in code modification tasks

Yi-Coder excels at cross-file code completion in CrossCodeEval

In the 'Needle in the code' task, Yi-Coder -9B demonstrates its long context modeling capabilities

In the Program-Aid Math Reasoning assessment, Yi-Coder -9B performed well in mathematical reasoning

Usage tutorial:

Visit Yi-Coder ’s GitHub page for model details and usage guidelines

Download and install necessary software dependencies, such as the Python environment and Transformers library

Set the environment for model training or inference according to Yi-Coder 's README document

Use the Yi-Coder API for code generation or join existing code editing tasks

Integrate Yi-Coder into the project and use its long context understanding and code generation capabilities to optimize the development process

Participate in Yi-Coder ’s community discussions to obtain technical support and best practice sharing

Contact Yi-Coder ’s development team via Discord or email for more in-depth help and discussion