So far, we have successfully built the machine learning model using Python and Sckit-Learn and a simple API using Flask. In this lesson, we will discuss alternative approaches, technologies, and frameworks you can use to do all we have built. During the workshop, we will discuss some of the use cases and the pros/cons.
| Framework | Creator(s) | Description |
|---|---|---|
| Tensorflow | An end-to-end machine learning platform for creating production-grade machine learning models built with C++ and Python. It has a comprehensive set of tools, libraries, and pre-trained models you can use to build machine-learning-powered applications. Some of the notable Tensorflow libraries are: TensorFlow.js—for creating machine learning models with JavaScript; Tensorflow Lite—for mobile and embedded devices like Android, iOS, Edge TPU, and Raspberry Pi, and Tensorflow TFX—for deploying production ML pipelines. | |
| Pytorch | Meta AI | A machine learning framework for building deep learning models on irregular input data built with Python and C++. It has a comprehensive set of tools and libraries, and tutorials you can use to build machine-learning-powered applications. |
| Keras | A neural-network deep-learning API designed for human beings that offers consistent and simple APIs built using Python. It has a comprehensive set of guides and examples you can use to build machine-learning-powered applications. | |
| Apache MXNet | Apache | A deep learning framework designed for both efficiency and flexibility for deep neural networks built using C++ and Python. It has a comprehensive set of tutorials and libraries you can use to build machine-learning-powered applications. |
| The Open Neural Network Exchange (ONNX) | Meta and Microsoft | An open ecosystem that empowers AI developers to choose the right tools as their project evolves with open source format for AI models (both deep learning and traditional ML). It has a comprehensive set of tutorials, Pre-trained ONNX models and framework/converters you can use to build machine-learning-powered applications. |
| Library | Description |
|---|---|
Joblib (.joblib) |
Like Pickle, this is used to serilize Python objects (mostly Sckit-Learn models). |
Hierarchical Data Format 5—HDF5 (.h5) |
Mostly used for Tensorflow models. |
Keras v3 Format (.keras) |
Mostly used for Keras models. |
| Keras SavedModel Format | Mostly used for Keras models. |
| JSON/YAML | Mostly used for Keras models. |
| Framework | Description |
|---|---|
| Django | A high-level Python web framework that encourages rapid development and clean, pragmatic design. |
| FastAPI | A modern, fast, web framework for building APIs with Python 3.7+. |
| Starlette | A lightweight ASGI framework/toolkit for building async web services in Python. |
| Sanic | A Python 3.8+ web server and web framework that's written to go fast. |
| Tornado | A Python web framework and asynchronous networking library. |
| Bottle | A fast, simple and lightweight WSGI micro web-framework for Python. |
| Pyramid | A small, fast, down-to-earth, open source Python web framework. |
Thank you for coming this far; you've done well 👏🏾. Please open a new GitHub discussion using the links below and let me know your thoughts about this lesson or any issues you're experiencing.
<< previous lesson | next lesson >>