ematvey/hierarchical-attention-networks
TensorFlow implementation of Hierarchical Attention Networks for document classification based on the 2016 Yang et al. paper.
Not currently ranked — collecting fresh signals.
star history
This repository implements a hierarchical attention network architecture for text classification, as described in the 2016 paper by Yang et al. The model uses word-level and sentence-level attention mechanisms to weight important content when classifying documents. It is built with TensorFlow and was trained on the Yelp review dataset, achieving 65% accuracy on a dev set.
Frequently asked
- What is ematvey/hierarchical-attention-networks?
- TensorFlow implementation of Hierarchical Attention Networks for document classification based on the 2016 Yang et al. paper.
- Is hierarchical-attention-networks open source?
- Yes — ematvey/hierarchical-attention-networks is open source, released under the MIT license.
- What language is hierarchical-attention-networks written in?
- ematvey/hierarchical-attention-networks is primarily written in Python.
- How popular is hierarchical-attention-networks?
- ematvey/hierarchical-attention-networks has 468 stars on GitHub.
- Where can I find hierarchical-attention-networks?
- ematvey/hierarchical-attention-networks is on GitHub at https://github.com/ematvey/hierarchical-attention-networks.