Add hasktorch to ideas for 2020

content/ideas/hasktorch.md

@@ -0,0 +1,54 @@
+---
+title: Hasktorch library for neural networks and tensor math
+---
+
+Hasktorch is a library for neural networks and tensor math in Haskell leveraging the C++ backend of PyTorch for fast numerical computation with GPU support. Our goal with Hasktorch is to provide a platform for machine learning using typed functional programming.
+
+As a summer project, there are a number of potential areas to contribute:
+
+### Write an in-depth tutorial series
+
+Port a tutorial series on neural networks such as tensor2tensor or fastai. Alongside code, write detailed comments/notes as well as adding to/improving//debugging the library when needed. 
+
+Relevant Links:
+
+- [FastAI](https://course.fast.ai/) - Practical Deep Learning for Coders course
+- [Trax](https://github.com/google/trax) - Example neural network implementations from Google Brain
+
+
+### PyTorch Interop
+
+Implement saving/loading serialized representations of models so that models can be transferred between PyTorch and Hasktorch.
+
+Relevant Links:
+
+- [torchscript](https://pytorch.org/docs/stable/jit.html) - an intermediate representation target for PyTorch models
+- [Hasktorch PR 268](https://github.com/hasktorch/hasktorch/pull/268) supporting torchscript types
+
+### Expand ecosystem around higher-level functionality
+
+Implement/improve modules for data loading, vision/text libraries, model visualization and interpretability, probabilistic modeling, etc.
+
+Relevant Links:
+
+- [torchvision](https://pytorch.org/docs/stable/torchvision/index.html) and [torchtext](https://pytorch.org/text/) documentation
+- [tf lucid](https://github.com/tensorflow/lucid) - a collection of infrastructure and tools for research in neural network interpretability.
+
+### Contribute to Foundational Low-level Code
+
+Help refine/debug foundational C++ foreign-function interface and code generation implementation. Improve resource management, help with migrations tracking upstream libtorch/PyTorch releases.
+
+Relevant Links:
+
+- [Hasktorch wiki docs on FFI](https://github.com/hasktorch/hasktorch/wiki/FFI-bindings-for-libtorch-library)
+- [Hasktorch wiki docs on resource management checking](https://github.com/hasktorch/hasktorch/wiki/How-to-show-resource-usage)
+
+## Potential Mentors:
+
+- Junji Hashimoto
+- Austin Huang
+- Adam Paszke
+- Torsten Scholak
+- Sam Stites
+
+**Difficulty**: Intermediate