蚂蚁 inclusionAI 发布 Ling Cookbook：Ling 模型系列的部署指南与实战示例
GitHub - inclusionAI/ling-cookbook: Guide, document and practical tips for getting the most out of the Ling model series. · GitHub
蚂蚁 inclusionAI 在 GitHub 上线 ling-cookbook，为 Ling 模型系列提供部署配方与工程实践示例，目前仍在完善中。仓库包含面向消费级与工作站硬件的本地推理部署指南，以及部分模型的多模态与智能体 Agent Skills。NVIDIA 团队贡献了 SGLang Humming 部署指南，在 Grace Blackwell（GB10 / SM121）上实现 Humming MoE 算子加速、在线 FP8 LM Head 与 MTP 推测解码协同。
正文
当前显示原文，中文全文译文尚未就绪。
inclusionAI; ling-cookbook
Repository navigation
NotificationsYou must be signed in to change notification settings; Fork7 (7); Star30 (30)You must be signed in to star a repository
Guide, document and practical tips for getting the most out of the Ling model series.
Repository files navigation
README
English | 简体中文
Ling Cookbook provides deployment recipes and practical engineering examples for the Ling model family. Work in progress.
Official Website: ant-ling.com; Hugging Face: inclusionAI; ModelScope: inclusionAI; GitHub: inclusionAI/ling-cookbook; Discord: Ling Community; X (Twitter): @AntLingAGI
Table of Contents
Local Deployment (guide/local-deploy/)
Recipes for deploying Ling models for local inference across consumer and workstation hardware.
Agent Skills for Specialized Capabilities
Some of our models provide distinctive multimodal and agentic capabilities. To help users explore these features, we provide pre-built Agent Skills. You can download and use them locally, or clone the repository to let your agent inspect the skills, audit their safety, and assist you in experiencing the model capabilities.
These agent skills and their associated models include:
English | 简体中文
Building Documentation
This project uses Jupytext to maintain all Jupyter Notebooks. All deployment guides and examples are authored in src/ (MyST Markdown) and compiled to their delivery destinations via scripts/compile_notebooks.py (*-zh.md to *-zh.ipynb, *.md to *.ipynb).
After editing files under src/, run python3 scripts/compile_notebooks.py to compile the notebooks. Direct manual edits to .ipynb files are prohibited.
Acknowledgements
Special thanks to the NVIDIA team @ly01325 for optimizing and contributing the SGLang Humming Deployment Guide, delivering Humming MoE operator acceleration, online FP8 LM Head, and MTP speculative decoding synergy on Grace Blackwell (GB10 / SM121) hardware
Guide, document and practical tips for getting the most out of the Ling model series.