Embeddings with Chromadb

Embeddings have revolutionized the way we represent and process data in machine learning and natural language processing tasks. They allow us to transform high-dimensional, unstructured data into compact, meaningful vector representations. ChromaDB, a powerful vector database, takes embeddings to the next level by providing efficient storage, retrieval, and similarity search capabilities. In this blog post, we'll explore how ChromaDB empowers developers to harness the full potential of embeddings.

What is embeddings?

Setup Chroma DB

Install chromadb package

pip install chromadb

Get the chroma client

import chromadb
chroma_client = chromadb.Client()

Create a collection

Collections are where you'll store your embeddings, documents, and any additional metadata. You can create a collection with a name:

collection = chroma_client.create_collection(name="my_collection")

Add some text documents to the collection

Chroma will store your text and handle embedding and indexing automatically. You can also customize the embedding model.

    collection.add(
        documents=[
            "This is a document about pineapple",
            "This is a document about oranges"
        ],
        ids=["id1", "id2"]
    )

Query the collection

results = collection.query(
    query_texts=["This is a query document about hawaii"], # Chroma will embed this for you
    n_results=2 # how many results to return
)
print(results)

Inspect Results

Links

• ChromaDB: Getting Started