DFL

A compiler of a Differentiable Functional Language.

I am providing code in the repository to you under an open source license.
Because this is my personal repository, the license you receive to my code is
from me and not my employer (Facebook).

What is it about?

The goal of this project is to explore relation between programs and
structures of neural networks.

There clearly is a correspondence between execution of a program, written in
functional paradigm, and flow of data in a neural network. I’ve already applied
this idea to design of sophisticated networks (see the FNN
language), which could learn non-trivial algorithms.

This repository is a next iteration of the research. I’ll try to implement a
compiler of a differentiable functional language with a haskell-like syntax.

How to use it?

Currently the project is under development, but one can check it out, build and
test:

stack build     # Build the project
stack test      # Run tests
stack haddock   # Build documentation
stack run my_file.dfl         # Run the compiler
stack run my_file.dfl -- -d   # Run the compiler and produce
                              # debug output for every step of
                              # compilation

License

This project is licensed under the MIT License - see the LICENSE file for details.