liveBook · Manning

Chroma DB

As explained in Data Storytelling with Altair and AI

Chroma is an embedding database for storing and retrieving vector representations of data.

Chroma is an embedding database that stores vector representations and supplies relevant context to language-model applications. In retrieval-augmented generation, it connects product information to question answering. section 7.4.2

Product-description retrieval

A persisted Product-Info collection retrieves chunks from product.html to answer a request describing the New SmartX 2023 in 30 words, returning both an answer and source documents.

How Data Storytelling with Altair and AI covers it

How deep it goes
Explains loading, chunking, embedding, persistence, retrieval, and output, without detailing Chroma's internal indexing.
Examples
Uses an HTML smartphone description, a Product-Info collection, and a generated SmartX 2023 answer with source documents.
What it assumes
Assumes familiarity with Python-based application code while introducing document loading, text splitting, embeddings, and retrieval progressively.
What you can do after
Build it from scratch by persisting embedded documents in Chroma and querying them through a LangChain retrieval application.
How it connects
Connects embeddings, vector databases, LangChain, retrieval-augmented generation, prompt templates, and large language models.

Also explained in

Compare how 7 books explain Chroma DB →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Data Storytelling with Altair and AI. The section links point to the parts of the book it draws on.