Bayesian Compression for Natural Language Processing

P. 2910–2915.

Chirkova N., Lobacheva E., Vetrov D.

In natural language processing, a lot of the tasks are successfully solved with recurrent neural networks, but such models have a huge number of parameters. The majority of these parameters are often concentrated in the embedding layer, which size grows proportionally to the vocabulary length. We propose a Bayesian sparsification technique for RNNs which allows compressing the RNN dozens or hundreds of times without time-consuming hyperparameters tuning. We also generalize the model for vocabulary sparsification to filter out unnecessary words and compress the RNN even further. We show that the choice of the kept words is interpretable. 

Keywords: автоматическая обработка естественного языка, deep learning, глубинное обучение, байесовские методы, Bayesian methods, Neural Language Processing (NLP), neural networks sparsification, сжатие нейронных сетей

In book

Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing

Association for Computational Linguistics, 2018.