Syntactic realization with data-driven neural tree grammars

Brian McMahan and Matthew Stone, 2016
Presented at COLING, 2016
Paper link (from repo)

important

Until I update this more fully, please be aware:

1. This uses my dev branch of keras
2. It also uses two helper libraries: baal and ikelos

short blurb

This work presents two stochastic tree models.
It evaluates them on supertag classification and string linearization.

This readme will be completed more in depth, soon.
Until then, check out scripts or runners.
They serve as the entry point to the training, the tagging, and the linearization.

README goals:

• description of each part of the repo.
• instructions on how to run