ml-quick-api 0.1.1

Turn a Model into an API in One Line

Quick-API

Turn a Model into an API in One Line

Quick-API is a Python library that wraps saved machine learning models (like .pkl, .h5, .joblib files) in a simple REST API using FastAPI with just one line of code.

Features

• 🚀 One-line API creation - Turn any saved model into a REST API instantly
• 🔄 Auto data conversion - Automatically handles JSON to NumPy array conversion
• 📊 Multiple model formats - Supports scikit-learn, TensorFlow/Keras, PyTorch models
• ⚡ FastAPI powered - Built on FastAPI for high performance and automatic documentation
• 🔍 Automatic endpoint discovery - Creates /predict endpoint automatically
• 📝 Interactive docs - Get Swagger UI documentation out of the box
• 🛡️ Input validation - Built-in request validation and error handling

Installation

pip install ml-quick-api

Quick Start

1. Basic Usage

from quick_api import create_api

# Turn your model into an API with one line
api = create_api("path/to/your/model.pkl")

# Run the API
api.run()

2. Advanced Usage

from quick_api import create_api

# Create API with custom configuration
api = create_api(
    model_path="models/my_classifier.pkl",
    host="0.0.0.0",
    port=8080,
    title="My ML API",
    description="A custom machine learning API",
    version="1.0.0"
)

# Run with custom settings
api.run(reload=True, workers=4)

3. Using the API

Once your API is running, you can make predictions:

curl -X POST "http://localhost:8000/predict" \
     -H "Content-Type: application/json" \
     -d '{"data": [[1.0, 2.0, 3.0, 4.0]]}'

Or visit http://localhost:8000/docs for interactive Swagger documentation.

Supported Model Types

• Scikit-learn models (.pkl, .joblib)
• TensorFlow/Keras models (.h5, .keras, saved_model format)
• PyTorch models (.pt, .pth)
• Custom models with predict method

API Endpoints

Endpoint	Method	Description
/predict	POST	Make predictions with your model
/health	GET	Check API health status
/info	GET	Get model information
/docs	GET	Interactive API documentation

Examples

Scikit-learn Example

import joblib
from sklearn.ensemble import RandomForestClassifier
from sklearn.datasets import make_classification
from quick_api import create_api

# Train and save a model
X, y = make_classification(n_samples=1000, n_features=4)
model = RandomForestClassifier()
model.fit(X, y)
joblib.dump(model, "classifier.pkl")

# Create API
api = create_api("classifier.pkl")
api.run()

TensorFlow Example

import tensorflow as tf
from quick_api import create_api

# Assuming you have a saved TensorFlow model
api = create_api("path/to/model.h5")
api.run()

Configuration Options

api = create_api(
    model_path="model.pkl",           # Path to your model file
    host="localhost",                 # Host to run the API on
    port=8000,                       # Port to run the API on
    title="Quick-API",               # API title
    description="ML Model API",      # API description
    version="1.0.0",                # API version
    input_shape=None,                # Expected input shape (auto-detected)
    preprocess_func=None,            # Custom preprocessing function
    postprocess_func=None,           # Custom postprocessing function
)

CLI Usage

Quick-API also provides a command-line interface:

# Basic usage
quick-api serve model.pkl

# With custom options
quick-api serve model.pkl --host 0.0.0.0 --port 8080 --title "My API"

Development

# Clone the repository
git clone https://github.com/yourusername/quick-api.git
cd quick-api

# Install in development mode
pip install -e ".[dev]"

# Run tests
pytest

# Format code
black quick_api/

# Type checking
mypy quick_api/

Contributing

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

License

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

Changelog

v0.1.0

• Initial release
• Support for scikit-learn, TensorFlow, and PyTorch models
• FastAPI-based REST API
• Automatic documentation
• CLI interface