Package profile
autograd
- Summary: Efficiently computes derivatives of NumPy code.
- Author: Dougal Maclaurin <maclaurin@physics.harvard.edu>, David Duvenaud <duvenaud@cs.toronto.edu>, Matthew Johnson <mattjj@csail.mit.edu>, Jamie Townsend <j.h.n.townsend@uva.nl>
- License: MIT
- Source: https://github.com/HIPS/autograd
- Number of releases: 33
- First release: 1.0.1 on 2015-04-10
- Latest release: 1.9.1 on 2026-07-02
- Latest release size: 53.1 KB (pure Python wheel)
Dependencies
Autograd has 2 dependencies, 1 of which optional.View all 2 dependencies →Dependent packages
| Package | Optional | Group |
|---|---|---|
| lifelines | false | |
| lifetimes | false | |
| pymoo | false | |
| pysisyphus | false | |
| nevergrad | true | all |
Similar packages
- numexprFast numerical expression evaluator for NumPy
- jaxDifferentiate, compile, and transform Numpy code.
- bottleneckFast NumPy array functions written in C
- nevergradA Python toolbox for performing gradient-free optimization
- numdifftoolsSolves automatic numerical differentiation problems in one or more variables.
- uprootROOT I/O in pure Python and NumPy.
- nptypingType hints for NumPy.