Release stack-augmented network architecture.

[adrelino]

Will you release the updated network from your new arxiv submission https://arxiv.org/abs/1912.11393?

This paper extends our work that first appeared in [6], adding analysis on reward shaping and the dependence on training set size, as well as the stack-augmented network architecture. Our PyTorch [7] implementation is publicly available at: https://hippogriff.github.io/CSGNet/.

[Hippogriff]

Hi! Thank you for your interest. The implementation of the new architecture will be released after the paper has been accepted.