google/jaxopt
A Google library providing hardware-accelerated, vectorizable, and differentiable optimization algorithms built on JAX.
Not currently ranked — collecting fresh signals.
star history
JAXopt implements a collection of gradient-based optimization algorithms designed to run efficiently on GPU, TPU, and CPU. It leverages JAX’s vmap for automatic vectorization of multiple optimization problems and supports implicit differentiation for differentiating through optimization solutions. The library is primarily used for training deep learning models, bi-level optimization tasks, and research involving differentiable programming.
Frequently asked
- What is google/jaxopt?
- A Google library providing hardware-accelerated, vectorizable, and differentiable optimization algorithms built on JAX.
- Is jaxopt open source?
- Yes — google/jaxopt is open source, released under the Apache-2.0 license.
- What language is jaxopt written in?
- google/jaxopt is primarily written in Python.
- How popular is jaxopt?
- google/jaxopt has 1k stars on GitHub.
- Where can I find jaxopt?
- google/jaxopt is on GitHub at https://github.com/google/jaxopt.