LLM Learning Hub

workspace/llm-course/home

Document Loading

Document loading is the process of reading files (PDF, HTML, Markdown, DOCX) and converting them into a uniform text-document object that downstream RAG components can chunk and embed.

What is Document Loading?

Document loading is the process of reading files (PDF, HTML, Markdown, DOCX) and converting them into a uniform text-document object that downstream RAG components can chunk and embed.

Document loading is the process of reading files (PDF, HTML, Markdown, DOCX) and converting them into a uniform text-document object that downstream RAG components can chunk and embed.

Where is it used?

LangChain `PyPDFLoader`, `WebBaseLoader`, and LlamaIndex `SimpleDirectoryReader` handle loading; Unstructured.io supports 25+ file types for enterprise RAG pipelines.

How to build it

In LangChain, use `PyPDFLoader("doc.pdf").load()` to get a list of `Document(page_content=..., metadata={...})` objects; inspect `.page_content` and `.metadata` to verify extraction quality.