Papers
arxiv:2205.02340

Knowledge Distillation of Russian Language Models with Reduction of Vocabulary

Published on May 4, 2022
Authors:
,
,

Abstract

Today, transformer language models serve as a core component for majority of natural language processing tasks. Industrial application of such models requires minimization of computation time and memory footprint. Knowledge distillation is one of approaches to address this goal. Existing methods in this field are mainly focused on reducing the number of layers or dimension of embeddings/hidden representations. Alternative option is to reduce the number of tokens in vocabulary and therefore the embeddings matrix of the student model. The main problem with vocabulary minimization is mismatch between input sequences and output class distributions of a teacher and a student models. As a result, it is impossible to directly apply KL-based knowledge distillation. We propose two simple yet effective alignment techniques to make knowledge distillation to the students with reduced vocabulary. Evaluation of distilled models on a number of common benchmarks for Russian such as Russian SuperGLUE, SberQuAD, RuSentiment, ParaPhaser, Collection-3 demonstrated that our techniques allow to achieve compression from 17times to 49times, while maintaining quality of 1.7times compressed student with the full-sized vocabulary, but reduced number of Transformer layers only. We make our code and distilled models available.

Community

Sign up or log in to comment

Models citing this paper 5

Browse 5 models citing this paper

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2205.02340 in a dataset README.md to link it from this page.

Spaces citing this paper 1

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.