Straight forward. Simple. Powerful.
Three extra lines and your model goes online.
Pinferencia (python + inference) aims to provide the simplest way to serve any of your machine
learning models with a fully functioning Graphic UI and Rest API.
Any model, in any framework, or even a function. You can serve it with Pinferencia.
Minimal codes to write, minimum codes modifications needed.
Just based on what you have.
Both statement and branch coverages, no kidding.
End-to-end tests with playwright.
Both backend and frontend codes are tested more than 100%.
All in Python. Straight forward model registration.
Hot reload. Fully use python3 type hints.
Develop, serve and debug at the same time.
Just a python library. It's light.
Built on FastAPI, it's fast.
With built-in streamlit support, it's amazing.
Built-in templates made with streamlit give you the extrordinary user experience out of box. All API explained in details with online try-out feature. Thanks to FastAPI and Starlette.
Pinferencia is designed to require minimum codes from developers and provide a clear way to register models at the same time. Just look at the example on the right.
Perhaps now you are thinking: it looks familiar!
Yes. We know you're not a user with no programming skills. And we know you're a developer capable of writing codes to train, evaluate, and test. Staying in Python and writing some codes can't be more comfortable for you.
That's why we bring you Pinferencia.
Instead of using model files and config files, we let you serve your models programmatically in Python, which is a lot easier to debug.
Pinferencia provides a beautiful graphic user interface made with Streamlit with built-in templates.
Everyone loves Streamlit, so do we. If you can't find the appropriate template, you are able to customize your own template. With the power of Pinferencia and Streamlit, everyone will be impressed by your app.
Pinferencia is not just for demostration of your model. It also deploys your model with fully functioning REST API sets.
There are two API set currently implemented: Default API and Kserve API.
Once the service online, you can visit the backend address and find all the specifications of the REST API. You can even try out the APIs using the interactive documentation web page.
Learn more
Pinferencia consists of a frontend and a backend.
You can choose to start the frontend and backend together, or start only
the frontend or the backend.
This gives you the power to build a service mesh, or deploy the backend in the cloud and only start the frontend locally.
Learn more Pinferencia is a part-time project from a small team of friends, us. We are trying to make a
tool that is useful to others.
Therefore, we would really appreciate if you find anything you'd like to share with us,
suggestions, bugs, or simply some thoughts.