embed-rerank 1.1.1

Single Model Embedding & Reranker API with Apple Silicon acceleration

🔥 Single Model Embedding & Reranking API

Lightning-fast local embeddings & reranking for Apple Silicon (MLX-first, OpenAI & TEI compatible)

---

⚡ Why This Matters

Transform your text processing with 10x faster embeddings and reranking on Apple Silicon. Drop-in replacement for OpenAI API and Hugging Face TEI with zero code changes required.

🏆 Performance Comparison

Operation	This API (MLX)	OpenAI API	Hugging Face TEI
Embeddings	0.78ms	200ms+	15ms
Reranking	1.04ms	N/A	25ms
Model Loading	0.36s	N/A	3.2s
Cost	$0	$0.02/1K	$0

Tested on Apple M4 Max

---

🚀 Quick Start

Option 1: Install from PyPI (Recommended)

# Install the package
pip install embed-rerank

# Start the server (default port 9000)
embed-rerank

# Or with custom port and options
embed-rerank --port 8080 --host 127.0.0.1

# See all options
embed-rerank --help

Option 2: From Source (Development)

# 1. Clone and setup
git clone https://github.com/joonsoo-me/embed-rerank.git
cd embed-rerank
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

# 2. Start server (macOS/Linux)
./tools/server-run.sh

# 3. Test it works
curl http://localhost:9000/health/

🎉 Done! Visit http://localhost:9000/docs for interactive API documentation.

---

🛠 Server Management (macOS/Linux)

# Start server (background)
./tools/server-run.sh

# Start server (foreground/development)
./tools/server-run-foreground.sh

# Stop server
./tools/server-stop.sh

Windows Support: Coming soon! Currently optimized for macOS/Linux.

---

⚙️ CLI Configuration

PyPI Package CLI Options

Server Options:

• --host: Server host (default: 0.0.0.0)
• --port: Server port (default: 9000)
• --reload: Enable auto-reload for development
• --log-level: Set log level (DEBUG, INFO, WARNING, ERROR, CRITICAL)

Testing Options:

• --test quick: Run quick validation tests
• --test performance: Run performance benchmark tests
• --test quality: Run quality validation tests
• --test full: Run comprehensive test suite
• --test-url: Custom server URL for testing
• --test-output: Test output directory

Examples:

# Custom server configuration
embed-rerank --port 8080 --host 127.0.0.1 --reload

# Built-in performance testing
embed-rerank --port 8080 &
embed-rerank --test performance --test-url http://localhost:8080
pkill -f embed-rerank

# Environment variables
export PORT=8080 HOST=127.0.0.1
embed-rerank

Source Code Configuration

Create .env file for development:

# Server
PORT=9000
HOST=0.0.0.0

# Backend
BACKEND=auto                                   # auto | mlx | torch
MODEL_NAME=mlx-community/Qwen3-Embedding-4B-4bit-DWQ

# Model Cache (first run downloads ~2.3GB model)
MODEL_PATH=                               # Custom model directory
TRANSFORMERS_CACHE=                           # HF cache override
# Default: ~/.cache/huggingface/hub/

# Performance
BATCH_SIZE=32
MAX_TEXTS_PER_REQUEST=100

---

📂 Model Cache Management

The service automatically manages model downloads and caching:

Environment Variable	Purpose	Default
MODEL_PATH	Custom model directory	(uses HF cache)
TRANSFORMERS_CACHE	Override HF cache location	~/.cache/huggingface/transformers
HF_HOME	HF home directory	~/.cache/huggingface
(auto)	Default HF cache	~/.cache/huggingface/hub/

Cache Location Check

# Find where your model is cached
python3 -c "
import os
print('MODEL_PATH:', os.getenv('MODEL_PATH', '<not set>'))
print('TRANSFORMERS_CACHE:', os.getenv('TRANSFORMERS_CACHE', '<not set>'))
print('HF_HOME:', os.getenv('HF_HOME', '<not set>'))
print('Default cache:', os.path.expanduser('~/.cache/huggingface/hub'))
"

# List cached Qwen3 models
ls ~/.cache/huggingface/hub | grep -i qwen3 || echo "No Qwen3 models found in cache"

---

🌐 Three APIs, One Service

API	Endpoint	Use Case
Native	/api/v1/embed, /api/v1/rerank	New projects
OpenAI	/v1/embeddings	Existing OpenAI code
TEI	/embed, /rerank	Hugging Face TEI replacement

OpenAI Compatible (Drop-in)

import openai

client = openai.OpenAI(
    api_key="dummy-key",
    base_url="http://localhost:9000/v1"
)

response = client.embeddings.create(
    input=["Hello world", "Apple Silicon is fast!"],
    model="text-embedding-ada-002"
)
# 🚀 10x faster than OpenAI, same code!

TEI Compatible

curl -X POST "http://localhost:9000/embed" 
  -H "Content-Type: application/json" 
  -d '{"inputs": ["Hello world"], "truncate": true}'

Native API

# Embeddings
curl -X POST "http://localhost:9000/api/v1/embed/" 
  -H "Content-Type: application/json" 
  -d '{"texts": ["Apple Silicon", "MLX acceleration"]}'

# Reranking  
curl -X POST "http://localhost:9000/api/v1/rerank/" 
  -H "Content-Type: application/json" 
  -d '{"query": "machine learning", "passages": ["AI is cool", "Dogs are pets", "MLX is fast"]}'

---

🧪 Performance Testing & Validation

🚀 Built-in CLI Testing (PyPI Package)

The PyPI package includes powerful built-in testing capabilities:

# Quick validation (basic functionality check)
embed-rerank --test quick

# Performance benchmark (latency, throughput, concurrency)
embed-rerank --test performance --test-url http://localhost:9000

# Quality validation (semantic similarity, multilingual)  
embed-rerank --test quality --test-url http://localhost:9000

# Full comprehensive test suite
embed-rerank --test full --test-url http://localhost:9000

Test Results Include:

• 📊 Latency Metrics: Mean, P95, P99 response times
• 🚀 Throughput Analysis: Texts/sec processing rates
• 🔄 Concurrency Testing: Multi-threaded request handling
• 🧠 Semantic Validation: Quality of embeddings and reranking
• 🌍 Multilingual Support: Cross-language performance
• 📈 JSON Reports: Detailed metrics for automation

Example Output:

🧪 Running Embed-Rerank Test Suite
📍 Target URL: http://localhost:9000
🎯 Test Mode: performance

⚡ Performance Results:
• Latency: 0.8ms avg, 1.2ms max
• Throughput: 1,250 texts/sec peak  
• Concurrency: 5/5 successful (100%)
📁 Results saved to: ./test-results/performance_test_results.json

🔧 Advanced Testing (Source Code)

### 🔧 Advanced Testing (Source Code)

For development and comprehensive testing with the source code:

```bash
# Comprehensive test suite (shell script)
./tools/server-tests.sh

# Run with specific test modes
./tools/server-tests.sh --quick            # Quick validation only
./tools/server-tests.sh --performance      # Performance tests only
./tools/server-tests.sh --full             # Full test suite

# Custom server URL
./tools/server-tests.sh --url http://localhost:8080

# Manual health check
curl http://localhost:9000/health/

# Unit tests with pytest
pytest tests/ -v

---

🛠 Development & Deployment

Local Development (Source Code)

# Start server (background)
./tools/server-run.sh

# Start server (foreground/development)
./tools/server-run-foreground.sh

# Stop server
./tools/server-stop.sh

Production Deployment (PyPI Package)

# Install and run
pip install embed-rerank
embed-rerank --port 9000 --host 0.0.0.0

# With custom configuration
embed-rerank --port 8080 --reload --log-level DEBUG

# Background deployment
embed-rerank --port 9000 &

Windows Support: Coming soon! Currently optimized for macOS/Linux.

---

## 🚀 What You Get

### 🎯 Core Features
- ✅ **Zero Code Changes**: Drop-in replacement for OpenAI API and TEI
- ⚡ **10x Performance**: Apple MLX acceleration on Apple Silicon  
- 💰 **Zero Costs**: No API fees, runs locally
- 🔒 **Privacy**: Your data never leaves your machine
- 🎯 **Three APIs**: Native, OpenAI, and TEI compatibility
- 📊 **Production Ready**: Health checks, monitoring, structured logging

### 🧪 Built-in Testing & Benchmarking
- 📈 **CLI Performance Testing**: One-command benchmarking
- 🔄 **Concurrency Testing**: Multi-threaded request validation
- 🧠 **Quality Validation**: Semantic similarity and multilingual testing
- 📊 **JSON Reports**: Automated performance monitoring
- 🚀 **Real-time Metrics**: Latency, throughput, and success rates

### 🛠 Deployment Options
- 📦 **PyPI Package**: `pip install embed-rerank` for instant deployment
- 🔧 **Source Code**: Full development environment with advanced tooling
- 🌐 **Multi-API Support**: OpenAI, TEI, and native endpoints
- ⚙️ **Flexible Configuration**: Environment variables, CLI args, .env files

---

## � Quick Reference

### Installation & Startup
```bash
# PyPI Package (Production)
pip install embed-rerank && embed-rerank

# Source Code (Development)  
git clone https://github.com/joonsoo-me/embed-rerank.git
cd embed-rerank && ./tools/server-run.sh

Performance Testing

# One-command benchmark
embed-rerank --test performance --test-url http://localhost:9000

# Comprehensive testing
./tools/server-tests.sh --full

API Endpoints

• Native: POST /api/v1/embed/ and /api/v1/rerank/
• OpenAI: POST /v1/embeddings (drop-in replacement)
• TEI: POST /embed and /rerank (Hugging Face compatible)
• Health: GET /health/ (monitoring and diagnostics)

---

�📄 License

MIT License - build amazing things with this code!