15.3K ·GitHub
What is this?
This is a curated Chinese-language knowledge base of large language model (LLM) concepts and interview questions, aimed at algorithm/application engineers preparing for LLM-related job interviews.
This is a curated Chinese-language knowledge base of large language model (LLM) concepts and interview questions, aimed at algorithm/application engineers preparing for LLM-related job interviews.Is this a single book or a collection of separate topic notes?It is organized as a numbered directory of markdown notes (01 through 10, plus courses and references) covering everything from NLP basics to RLHF, deployed as an online reading site rather than a single linear book.
It is organized as a numbered directory of markdown notes (01 through 10, plus courses and references) covering everything from NLP basics to RLHF, deployed as an online reading site rather than a single linear book.Where should someone start reading?Start at the table of contents, which begins with real interview questions and chapter 01 on LLM fundamentals, then progresses through architecture, training, fine-tuning, inference, RLHF, RAG, and LangChain.
Start at the table of contents, which begins with real interview questions and chapter 01 on LLM fundamentals, then progresses through architecture, training, fine-tuning, inference, RLHF, RAG, and LangChain.Are the answers reviewed by experts?No, the answers are written by the single author based on online research, and the repo explicitly invites readers to point out and correct any inaccuracies.
No, the answers are written by the single author based on online research, and the repo explicitly invites readers to point out and correct any inaccuracies.Who is this guide meant for?It targets engineers and job candidates studying for LLM algorithm or application engineer interviews, covering topics from basic NLP to distributed training and RLHF.
It targets engineers and job candidates studying for LLM algorithm or application engineer interviews, covering topics from basic NLP to distributed training and RLHF.Are there hands-on companion projects to practice with?Yes, the author links several practice repos: tiny-llm-zh (building a small Chinese LLM from scratch, including pretraining, fine-tuning and RL), tiny-rag (a simple RAG system with multi-path retrieval and reranking), tiny-mcp (an MCP server/client built with prompts and function calling), and llama3-from-scratch-zh (a from-scratch Llama 3 implementation loadable with Meta's official weights, runnable on a 16GB laptop).
Yes, the author links several practice repos: tiny-llm-zh (building a small Chinese LLM from scratch, including pretraining, fine-tuning and RL), tiny-rag (a simple RAG system with multi-path retrieval and reranking), tiny-mcp (an MCP server/client built with prompts and function calling), and llama3-from-scratch-zh (a from-scratch Llama 3 implementation loadable with Meta's official weights, runnable on a 16GB laptop).Is all the content in Chinese?Yes, the notes, interview questions, and the linked WeChat account content are all written in Chinese, with no English version mentioned.
Yes, the notes, interview questions, and the linked WeChat account content are all written in Chinese, with no English version mentioned.What topics does the guide cover?It spans LLM basics, tokenization and embeddings, Transformer/attention/MoE architecture, distributed training (data/pipeline/tensor parallelism, DeepSpeed, Megatron), supervised fine-tuning (LoRA), inference and quantization, RLHF (PPO/DPO), RAG, agents, model evaluation/hallucination, and LangChain.
It spans LLM basics, tokenization and embeddings, Transformer/attention/MoE architecture, distributed training (data/pipeline/tensor parallelism, DeepSpeed, Megatron), supervised fine-tuning (LoRA), inference and quantization, RLHF (PPO/DPO), RAG, agents, model evaluation/hallucination, and LangChain.What are similar projects?ai-system-design, nlp_course, ai-engineering-interview-questions-company-wise, autumn-recruitment, AgentGuide, LLM101n, llm-course, DeepLearning
ai-system-design, nlp_course, ai-engineering-interview-questions-company-wise, autumn-recruitment, AgentGuide, LLM101n, llm-course, DeepLearningWhich topics is it in?interview-prep, rlhf, function-calling, coding-interview, distributed-training, mixture-of-experts, llama, llms, study-guide, interview-questions, natural-language-processing, langchain
interview-prep, rlhf, function-calling, coding-interview, distributed-training, mixture-of-experts, llama, llms, study-guide, interview-questions, natural-language-processing, langchain