port:

Visual interface for functions, machine learning models and APIs

Port is an experimental JavaScript library that simplifies porting algorithms to web browsers. It's based on the idea of declarative interface design and reactivity. Instead of writing "glue" front-end code, you can just declare inputs/outputs of a model in a JSON schema and Port will do the rest. Port creates input and output elements, parses files, loads needed libraries and uses web workers to run everything smoothly. It's not a swiss-army knife, not a framework. Port solves one specific task - wrapping algorithms in a web interface.

Javascript

Port regular JavaScript functions and classes or use async methods

TensorFlow

Import trained neural net models and run them with TensorFlow.js

Python

Run Python with its scientific stack in the browser with Pyodide.js

API

Test how your REST API works with a simple graphical UI

    JavascriptFactorial JavascriptQR Code generator JavascriptSentiment analysis PythonPandas summary