search

mlajtos / moniel

Interactive Notation for Computational Graphs
https://mlajtos.github.io/moniel/
MIT License
356 stars 27 forks source link

Add support for PyTorch #8

Open mlajtos opened 7 years ago

mlajtos commented 7 years ago

Generating executable code is a top priority. PyTorch is currently the top DL framework (because of the define-by-run nature), so naturally it would be very suiting to combine the two.

Moniel should provide easy way to define weird architectures and PyTorch is cabable of representing them. However, user must be able to interfere with the generated code, because high-level description often does not capture all the neccesities of the idea.

mlajtos commented 7 years ago

Since code is the best representation for computational graphs, output should be generated Python code using PyTorch infrastructure. Again, output WILL NOT be any JSON/YAML/PB stuff, but actual runnable code.

Since user will be able to change generated code (or maybe inject it from Moniel?), result will be the best of the both worlds.

mlajtos commented 7 years ago

There is a proof-of-concept code that constructs code that resembles PyTorch, but there are a few questions that need to be discussed: