AiHubCN/Awesome-Chinese-LLM

What it is

Awesome-Chinese-LLM is a curated, community-maintained list (an "awesome list") of open-source resources related to Chinese large language models (LLMs). It collects and organizes links to models, applications, datasets, and tutorials — over 100 resources at the time of writing.

Why it exists

The project was created in response to the rapid growth of LLMs after ChatGPT, especially the release of smaller open-source models like ChatGLM and LLaMA that individuals can run on consumer hardware. This sparked many fine-tuned variants and applications, and the list aims to make these discoverable.

What it contains

The README organizes resources into these main sections:

• Models — split into text-only LLMs (e.g., ChatGLM, Qwen, Baichuan, InternLM, DeepSeek, Yi, and many fine-tuned variants) and multimodal LLMs that handle images plus text (e.g., VisualGLM-6B, CogVLM).
• Applications — including vertical-domain fine-tuned models (medical, legal, financial, educational, etc.), LangChain-based apps, and other uses.
• Datasets — pretraining, supervised fine-tuning (SFT), and preference datasets.
• Training/fine-tuning frameworks and inference/deployment frameworks.
• Evaluation tools and tutorials (basics, prompt engineering, applications, hands-on practice).
• Related repositories.

Key feature: base model comparison table

A prominent table compares common base models (ChatGLM, LLaMA, Baichuan, Qwen, BLOOM, Aquila, InternLM, Mixtral, Yi, DeepSeek, XVERSE, MiniMax) across:

• included model variants
• parameter sizes
• training token counts
• maximum context length
• whether commercial use is allowed

Most listed models are marked as commercially usable, with LLaMA noted as "partially" commercial.

How to use it

Each entry provides a GitHub link, a star badge, and a short description in Chinese. The list is meant as a reference hub — readers browse to find relevant models or tools, then click through to the original repositories.

Contributing

The project explicitly invites contributions: users can submit pull requests to add new open-source models, applications, or datasets, following the existing format (repository link, star count, brief description).

Maintenance signals

The README includes GitHub stars, issues, and forks badges, plus a "Star History" section, indicating an actively tracked community project. The content is bilingual (Chinese primary, with some English labels).

In short: it's a well-organized directory of Chinese LLM resources, useful for researchers, developers, and hobbyists looking to find or compare open-source Chinese-language AI models and related tooling.