@@ -6,21 +6,24 @@ author: Huw Campbell <huw.campbell@gmail.com>
66maintainer: Huw Campbell <huw.campbell@gmail.com>
77copyright: (c) 2016-2017 Huw Campbell.
88synopsis: Practical Deep Learning in Haskell
9−category: System
9+category: AI, Machine Learning
1010cabal-version: >= 1.8
1111build-type: Simple
1212description:
1313 Grenade is a composable, dependently typed, practical, and fast
14− recurrent neural network library for concise and precise
15− specifications of complex networks in Haskell.
14+ recurrent neural network library for precise specifications and
15+ complex deep neural networks in Haskell.
1616 .
1717 Grenade provides an API for composing layers of a neural network
18− into a sequence parallel graph in a type safe manner.
18+ into a sequence parallel graph in a type safe manner; running
19+ networks with reverse automatic differentiation to calculate their
20+ gradients; and applying gradient decent for learning.
1921 .
2022 Documentation and examples are available on github
2123 <https://github.com/HuwCampbell/grenade>.
2224
2325extra-source-files:
26+ README.md
2427 cbits/im2col.h
2528 cbits/im2col.c
2629 cbits/gradient_decent.h