Flux: Elegant machine learning with Julia
By M. Innes. Journal of Open Source Software, 2018
382 citations recorded by the source corpus.
Flux is library for machine learning (ML), written using the numerical computing language Julia (Bezanson et al. 2017). The package allows models to be written using Julia’s simple mathematical syntax, and applies automatic differentiation (AD) to seamlessly calculate derivatives and train the model. Meanwhile, it makes heavy use of Julia’s language and compiler features to carry out code analysis and make optimisations. For example, Julia’s GPU compilation support (Besard, Foket, and De Sutter 2017) can be used to JIT-compile custom GPU kernels for model layers (Innes and others 2017a).
Subfields: Artificial Intelligence
View Paper: https://www.semanticscholar.org/paper/d4414002f23c0f1c497166eb51c5b1d549ff75c8
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Philipp Hennig
- Felix Dangel
- Frederik Kunstner
- Alexandre Passos
- Akshay Agrawal
- Ashish Agarwal
- Igor Ganichev
- Rajat Monga
Citations in the corpus
- BackPACK: Packing more into backprop (2019, background): “While other researchers are aiming to improve the flexibility of automatic differentiation systems (Innes, 2018a;b; Bradbury et al., 2018), our goal with this package is to provide access to quantities that are only byproducts of the backpropagation pass, rather than gradients themselves.”
- TensorFlow Eager: A Multi-Stage, Python-Embedded DSL for Machine Learning (2019, background): “…operations and functions to automatically differentiate compositions thereof (see, e.g., Bergstra et al., 2010; Tokui et al., 2015; Maclaurin et al., 2015; Chen et al., 2015; Abadi et al., 2016; Paszke et al., 2017; The Gluon Team, 2017; Neubig et al., 2017; Innes, 2018; Frostig et al., 2018).”