explosion/thinc
A lightweight deep learning library with a functional-programming API for composing models across multiple ML frameworks.
Not currently ranked — collecting fresh signals.
star history
Thinc is a Python deep learning library that provides a type-checked functional API for composing neural network layers. It works as an interface layer, allowing users to mix and match layer definitions from PyTorch, TensorFlow, MXNet, and JAX. Built by the makers of spaCy and Prodigy, it enables composing, configuring, and deploying custom models with a unified, elegant interface.
Frequently asked
- What is explosion/thinc?
- A lightweight deep learning library with a functional-programming API for composing models across multiple ML frameworks.
- Is thinc open source?
- Yes — explosion/thinc is open source, released under the MIT license.
- What language is thinc written in?
- explosion/thinc is primarily written in Python.
- How popular is thinc?
- explosion/thinc has 2.9k stars on GitHub.
- Where can I find thinc?
- explosion/thinc is on GitHub at https://github.com/explosion/thinc.