typedflow: Typed frontend to TensorFlow and higher-order deep learning

[ deep-learning, lgpl, library ] [ Propose Tags ]

TypedFlow is a typed, higher-order frontend to TensorFlow and a high-level library for deep-learning.

The main design principles are:

  • To make the parameters of layers explicit. This choice makes sharing of parameters explicit and allows to implement "layers" as pure functions.

  • To provide as precise as possible types. Functions are explicit about the shapes and elements of the tensors that they manipulate (they are often polymorphic in shapes and elements though.)

  • To let combinators be as transparent as possible. If a NN layers is a simple tensor transformation it will be exposed as such.

Downloads

Maintainer's Corner

Package maintainers

For package maintainers and hackage trustees

Candidates

Versions [RSS] 0.9
Dependencies base (>=4 && <5), ghc-typelits-knownnat, mtl, pretty-compact [details]
License LGPL-3.0-only
Author Jean-Philippe Bernardy
Maintainer jean-philippe.bernardy@gu.se
Category Deep Learning
Source repo head: git clone git@github.com:GU-CLASP/TypedFlow.git
Uploaded by JeanPhilippeBernardy at 2017-10-26T14:13:14Z
Distributions
Reverse Dependencies 1 direct, 0 indirect [details]
Downloads 974 total (6 in the last 30 days)
Rating (no votes yet) [estimated by Bayesian average]
Your Rating
  • λ
  • λ
  • λ
Status Docs available [build log]
Last success reported on 2017-10-26 [all 1 reports]