Chunking

In the context of building LLM-related applications, chunking is the process of breaking down large pieces of text into smaller segments. It’s an essential technique that helps optimize the relevance of the content we get back from a vector database once we use the LLM to embed content.

There are multiple aspects to the chunking processes such as:

• Chunk Size
• Chunking Strategies
• Metadata Addition
  • Enhancing chunk blocks with metadata like titles, keywords, and hypothetical questions can improve retrieval, provide more ways to post-process retrieved texts, and help LLMs better understand retrieved information.