uber/fiber
Fiber is a Python distributed computing library that enables AI model training across computer clusters using a multiprocessing-style API.
Not currently ranked — collecting fresh signals.
star history
Fiber provides a distributed computing framework that allows Python programs to run across cluster environments without requiring manual cluster management. It mimics Python’s standard multiprocessing API so developers familiar with multiprocessing can easily scale to cluster computing. The library uses Nanomsg for high-performance async communication and includes built-in fault tolerance for handling worker failures.
Frequently asked
- What is uber/fiber?
- Fiber is a Python distributed computing library that enables AI model training across computer clusters using a multiprocessing-style API.
- Is fiber open source?
- Yes — uber/fiber is open source, released under the Apache-2.0 license.
- What language is fiber written in?
- uber/fiber is primarily written in Python.
- How popular is fiber?
- uber/fiber has 1k stars on GitHub.
- Where can I find fiber?
- uber/fiber is on GitHub at https://github.com/uber/fiber.