Documentation / Quick Start

Quick Start.

Get up and running with Feather v0.5.0 in under 5 minutes. Learn how to initialize the DB, add multimodal vectors, and perform searches.

Step 1: Create a Database

Start by creating a new Feather database specifying the vector dimension:

Python

from feather import DB

# Create a database for 384-dimensional vectors
db = DB.open("embeddings.feather", dim=384)

Step 2: Add Multimodal Vectors

Add vectors to your database. In v0.5.0, you can specify individual modalities for the same entity ID:

Python

import numpy as np

vectors = [np.random.rand(384).astype(np.float32) for _ in range(10)]

for i, vector in enumerate(vectors):
    # Add with an explicit modality string
    db.add(id=i, vec=vector, modality="text")

print(f"Added {len(vectors)} text vectors")

Step 3: Search with Namespaces

Query the vector database. Results are sorted by closest distance (L2 or Cosine, depending on build params).

Python

query = np.random.rand(384).astype(np.float32)

# Search specifically within the "text" modality pockets
results = db.search(query, k=5, modality="text")

for vector_id, distance in results:
    print(f"ID: {vector_id}, Distance: {distance:.4f}")

Step 4: Save and Persist

Feather writes natively to a highly optimized flat binary file:

Python

# Flush index and data to disk
db.save()

# Load exactly the same way later:
# db = DB.open("embeddings.feather", dim=384)