Qdrant Vector Database Examples & Projects

This repository contains a collection of code examples and projects demonstrating the use of Qdrant, a powerful vector database for similarity search and machine learning applications.

Overview

Qdrant is a vector similarity search engine that provides a production-ready service with a convenient API to store, search, and manage points—vectors with an optional payload. This repository demonstrates various ways to use Qdrant for different use cases.

Prerequisites

• Python 3.7+
• Qdrant server running (default: localhost:6333)
• Required Python packages:
  • qdrant-client
  • sentence-transformers
  • transformers
  • torch
  • PIL (Pillow)
  • pandas
  • numpy

Installation

# Clone the repository
git clone https://github.com/yourusername/qdrant-samples.git
cd qdrant-samples

# Install dependencies
pip install -r requirements.txt

# Start Qdrant (if not already running)
# Using Docker:
docker run -p 6333:6333 qdrant/qdrant

Project Structure

qdrant-samples/
├── simple code/
│   └── qdrant_how_to_code.py  # Basic Qdrant operations
├── project/
│   ├── semantic_search.py     # Semantic search for text documents
│   ├── image_search.py        # Image similarity search
│   └── multilingual.py        # Multilingual text search
└── README.md

Projects

1. Basic Qdrant Operations (simple code/qdrant_how_to_code.py)

A simple introduction to Qdrant that covers:

• Creating a collection
• Inserting vectors with payloads
• Performing vector similarity search
• Displaying search results

This is an excellent starting point to understand the fundamental operations of Qdrant.

2. Semantic Search (project/semantic_search.py)

Demonstrates semantic search capabilities using:

• Sentence transformer models for text embeddings
• Document indexing with metadata
• Similarity search based on meaning rather than keywords
• Example with a collection of science fiction books

3. Image Search (project/image_search.py)

Implements image similarity search using:

• Vision Transformer (ViT) models for image embedding
• Image preprocessing and feature extraction
• Vector similarity for finding similar images
• Practical example with local image collection

4. Multilingual Search (project/multilingual.py)

Showcases cross-language semantic search:

• Multilingual embedding model supporting 50+ languages
• Cross-lingual similarity matching
• Filtering by language and topic
• Examples of querying in one language and finding relevant content in others

Usage Examples

Basic Vector Operations

from qdrant_client import QdrantClient
from qdrant_client.models import Distance, VectorParams, PointStruct

# Connect to Qdrant
client = QdrantClient("localhost", port=6333)

# Create a collection
client.create_collection(
    collection_name="my_collection",
    vectors_config=VectorParams(size=384, distance=Distance.COSINE)
)

# Insert vectors
client.upsert(
    collection_name="my_collection",
    points=[
        PointStruct(
            id=1,
            vector=[0.1, 0.2, ...],  # 384-dimensional vector
            payload={"metadata": "example"}
        )
    ]
)

# Search for similar vectors
results = client.search(
    collection_name="my_collection",
    query_vector=[0.1, 0.2, ...],  # Query vector
    limit=5
)

Semantic Text Search

# See project/semantic_search.py for complete implementation
from sentence_transformers import SentenceTransformer

# Initialize the model
model = SentenceTransformer('all-MiniLM-L6-v2')

# Create embeddings for your text
text_embedding = model.encode("Your search query").tolist()

# Search in Qdrant
results = client.search(
    collection_name="semantic_search",
    query_vector=text_embedding,
    limit=3
)

Image Similarity Search

# See project/image_search.py for complete implementation
from PIL import Image
from transformers import AutoFeatureExtractor, AutoModel

# Get embedding for query image
query_embedding = get_image_embedding("path/to/query_image.jpg")

# Search for similar images
results = client.search(
    collection_name="images",
    query_vector=query_embedding.tolist(),
    limit=5
)

Multilingual Search

# See project/multilingual.py for complete implementation
# Search for "artificial intelligence" in any language
results = search_documents("artificial intelligence")

# Search in specific language or topic
results = search_documents("énergie renouvelable", language="en", topic="energy")

Key Features

• Vector Similarity Search: Find semantically similar items using cosine distance
• Cross-Modal Applications: Text-to-text, image-to-image similarity
• Multilingual Support: Search across language barriers
• Metadata Filtering: Combine vector search with metadata constraints
• Efficient Indexing: Fast retrieval even with large vector collections

Applications

This repository demonstrates implementation patterns for:

• Semantic document search
• Similar image retrieval
• Multilingual information retrieval
• Product recommendation systems
• Content-based filtering

Resources

• Qdrant Documentation
• Sentence Transformers
• Hugging Face Transformers

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.