#74 · Primary category: Foundation Models

Baichuan2

artificial-intelligence benchmark ceval chatgpt chinese gpt gpt-4 huggingface large-language-models llama2 mmlu natural-language-processing

A series of large language models developed by Baichuan Intelligent Technology

Project last updated:11/08/24

GitHub Stars

4.1K

Forks

294

Contributors

10

License

Apache-2.0

Why we included this project

Baichuan 2 offers 7B and 13B open-weight language models from Baichuan Intelligent Technology, with base and chat versions for each size plus 4-bit quantized builds that keep memory requirements down. The chat models perform well on Chinese benchmarks like C-Eval and CMMLU, and Gaokao scores are competitive too, so teams building Chinese-language applications can self-host a capable model instead of calling an external API. The repository bundles inference notes, fine-tuning scripts, and intermediate checkpoints, letting you go from a downloaded model to a working service or a domain-adapted variant without assembling extra tooling. A permissive license with free commercial use on request makes it a low-friction choice for evaluation.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category