hmatrix-backprop: hmatrix operations lifted for backprop

[ bsd3, library, math ] [ Propose Tags ]

hmatrix operations lifted for backprop, along with orphan instances.

Meant to act as a drop-in replacement to the API of Numeric.LinearAlgebra.Static. Just change your imports, and your functions are automatically backpropagatable.

See README on Github at https://github.com/mstksg/hmatrix-backprop#readme


[Skip to Readme]
Versions 0.1.0.0, 0.1.1.0, 0.1.2.0, 0.1.2.1, 0.1.2.2, 0.1.2.3, 0.1.2.4, 0.1.2.5
Change log CHANGELOG.md
Dependencies backprop (>=0.2.5), base (>=4.7 && <5), ghc-typelits-knownnat, ghc-typelits-natnormalise, hmatrix (>=0.18), hmatrix-vector-sized (>=0.1.1), microlens, vector, vector-sized (>=0.6), vinyl (>=0.9) [details]
License BSD-3-Clause
Copyright (c) Justin Le 2018
Author Justin Le
Maintainer justin@jle.im
Category Math
Home page https://github.com/mstksg/hmatrix-backprop#readme
Bug tracker https://github.com/mstksg/hmatrix-backprop/issues
Source repo head: git clone https://github.com/mstksg/hmatrix-backprop
Uploaded by jle at Wed Nov 14 20:46:10 UTC 2018
Distributions LTSHaskell:0.1.2.3, NixOS:0.1.2.5, Stackage:0.1.2.5
Downloads 552 total (56 in the last 30 days)
Rating (no votes yet) [estimated by rule of succession]
Your Rating
  • λ
  • λ
  • λ
Status Docs not available [build log]
Last success reported on 2018-11-14 [all 3 reports]
Hackage Matrix CI

Modules

  • Numeric
    • LinearAlgebra
      • Static
        • Numeric.LinearAlgebra.Static.Backprop

Downloads

Maintainer's Corner

For package maintainers and hackage trustees


Readme for hmatrix-backprop-0.1.2.5

[back to package description]

hmatrix-backprop

hmatrix-backprop on Hackage Build Status

hmatrix operations lifted for backprop, along with orphan instances of Backprop.

Meant to act as a drop-in replacement to the API of Numeric.LinearAlgebra.Static. Just change your imports, and your functions are automatically backpropagatable. Useful types are all re-exported.

Formulas for gradients come from the following papers:

Some functions are not yet implemented! See module documentation for details. PR's definitely appreciated :)

Tests

Currently numeric tests are implemented as property tests using hedgehog, but it is possible that the answers might differ from the true values by an amount undetectable by property tests.

All functions currently are tested except for the higher-order functions.

They are tested by "nudging" components of inputs and checking if the change in the function outputs match what is expected from the backpropagated gradient.

TODO

  1. Now that new backprop no longer requires Num, we can lift normal hmatrix operations as well.
  2. Statically sized convolutions. Should probably add this to hmatrix instead first, though.