| Llama 大模型实践指南 | Llama Large Model Practice Guide |
Published in 电子工业出版社 | Publishing House of Electronics Industry, 2024
《Llama大模型实践指南》是国内首部系统介绍Llama大模型的中文实践指南，由Llama中文社区核心成员共同编写。内容涵盖Llama大模型架构解析、中文环境下的预训练与微调实践、32K长上下文支持与优化技巧等核心技术，以及企业级应用案例与部署方案。
The “Llama Large Model Practice Guide” is the first systematic Chinese guide to Llama large models, co-authored by core members of the Llama Chinese Community. It covers core technologies including Llama model architecture analysis, Chinese pretraining and fine-tuning practices, 32K long context support and optimization techniques, as well as enterprise application cases and deployment solutions.
作为Llama中文社区联合发起人，我负责书中关于中文预训练策略和LoRA高效微调方案的章节编写，分享了在Atom系列大模型研发中积累的核心技术经验。本书适合大模型研究人员、工程师以及对AI技术感兴趣的专业人士阅读。
As a co-founder of Llama Chinese Community, I authored chapters on Chinese pretraining strategies and LoRA fine-tuning, sharing core technical expertise from developing the Atom series models. This book is ideal for LLM researchers, engineers, and AI professionals.
Recommended citation: 张俊祺、曲东奇、张正、占冰强. (2024). "Llama 大模型实践指南" 电子工业出版社.
Download Paper