[PyCon Korea 2025] RAG 애플리케이션 개발을 위한 Chunking 최적화 - 강성우

Name: [PyCon Korea 2025] RAG 애플리케이션 개발을 위한 Chunking 최적화 - 강성우
Uploaded: 2025-12-22T00:00:00.000Z
Duration: 22 min 23 s
Description: Chunking is not a skippable “just split the text” step in RAG: Transformer LLMs have no built-in memory and a hard context length, so answers depend on how you cut documents, retrieve the right pieces, and restore con...

PyCon Korea · 22:23

Chunking is not a skippable “just split the text” step in RAG: Transformer LLMs have no built-in memory and a hard context length, so answers depend on how you cut documents, retrieve the right pieces, and restore con...

Read the full summary on tuber