HuwCampbell/grenade
A dependently typed Haskell library for building and specifying convolutional, recurrent, and generative adversarial neural networks.
Not currently ranked — collecting fresh signals.
star history
Grenade is a composable deep learning framework written in Haskell that allows concise, type-safe specifications of complex neural network architectures. It provides implementations for convolutional layers, pooling, LSTM cells, fully connected layers, and generative adversarial networks. The library leverages Haskell’s dependent typing to express network architectures and tensor shapes at the type level, ensuring compile-time correctness of layer dimensions.
Frequently asked
- What is HuwCampbell/grenade?
- A dependently typed Haskell library for building and specifying convolutional, recurrent, and generative adversarial neural networks.
- Is grenade open source?
- Yes — HuwCampbell/grenade is open source, released under the BSD-2-Clause license.
- What language is grenade written in?
- HuwCampbell/grenade is primarily written in Haskell.
- How popular is grenade?
- HuwCampbell/grenade has 1.4k stars on GitHub.
- Where can I find grenade?
- HuwCampbell/grenade is on GitHub at https://github.com/HuwCampbell/grenade.