hudson 0.0.0a1

Build multi-modal data applications with ease.

🎶If you havin' data problems I feel bad for you son.
I got 99 problems and a framework ain't one.🎶

---

:warning: Hudson is still in alpha and shouldn't be used for production yet. If you have any questions, feedback, or feature suggestions, please create an issue on Github.

Documentation: https://hudson.corletti.xyz

Source Code: https://github.com/anthonycorletti/hudson

---

Build multi-modal data applications with ease.

Some major features of Hudson are;

• 🐍 Async Python: Hudson is 100% async. It's built on top of FastAPI, SQLModel, Uvicorn, Pydantic, DocArray, and more.
• 🧱 DocArray: Hudson uses DocArray so you can work with multi-modal data without having to do work to support each modality separately.
• 🐻‍❄️ Polars: Hudson uses Polars for blazing fast server-side data processing.
• ☁️ Modal: Hudson deploys on Modal by default. No need to worry about infrastructure, Kubernetes, or containers!
• 📨 Publish-subscribe functionality built right in. Create any workflow you need with Hudson!
• ✍️ Just write code! No YAML necessary.

What's Hudson?

Hudson is a framework for building multi-modal data applications.

Hudson runs as a server-side application and provides a REST API for your application to communicate with. Hudson also provides a python client library that you can use to interact with the server.

Use cases

1. Multi-modal data analytics: One way to work with data across any modality.
2. Processing data in the cloud: Build data processing pipelines with ease – no need to worry about infrastructure, plugging different cloud tools together, or writing code to support each modality.
3. Machine learning data analytics: Send data from your machine learning workflow to Hudson for processing and analysis. It already works with any modality and framework – you just have to send in the data and embeddings and that's it!

Installing Hudson

Hudson is available on PyPI. You can install it with pip:

pip install hudson

Contributing & Sponsorship

One of the easiest and best ways to contribute to Hudson is to star the project on GitHub and share it with your colleagues, friends, and anyone who might want to build data applications without the hassle.

If you would like to contribute, please read Hudson's contributing guidelines. Issues and pull requests are very welcome.

If you would like to sponsor the project, you can do so here