Distilling Robustness into Natural Language Inference Models with Domain-Targeted Augmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Stacey, Joe, Rei, Marek
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913444081434624
author Stacey, Joe
Rei, Marek
author_facet Stacey, Joe
Rei, Marek
contents Knowledge distillation optimises a smaller student model to behave similarly to a larger teacher model, retaining some of the performance benefits. While this method can improve results on in-distribution examples, it does not necessarily generalise to out-of-distribution (OOD) settings. We investigate two complementary methods for improving the robustness of the resulting student models on OOD domains. The first approach augments the distillation with generated unlabelled examples that match the target distribution. The second method upsamples data points among the training set that are similar to the target distribution. When applied on the task of natural language inference (NLI), our experiments on MNLI show that distillation with these modifications outperforms previous robustness solutions. We also find that these methods improve performance on OOD domains even beyond the target domain.
format Preprint
id arxiv_https___arxiv_org_abs_2305_13067
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Distilling Robustness into Natural Language Inference Models with Domain-Targeted Augmentation
Stacey, Joe
Rei, Marek
Computation and Language
Machine Learning
I.2.7
Knowledge distillation optimises a smaller student model to behave similarly to a larger teacher model, retaining some of the performance benefits. While this method can improve results on in-distribution examples, it does not necessarily generalise to out-of-distribution (OOD) settings. We investigate two complementary methods for improving the robustness of the resulting student models on OOD domains. The first approach augments the distillation with generated unlabelled examples that match the target distribution. The second method upsamples data points among the training set that are similar to the target distribution. When applied on the task of natural language inference (NLI), our experiments on MNLI show that distillation with these modifications outperforms previous robustness solutions. We also find that these methods improve performance on OOD domains even beyond the target domain.
title Distilling Robustness into Natural Language Inference Models with Domain-Targeted Augmentation
topic Computation and Language
Machine Learning
I.2.7
url https://arxiv.org/abs/2305.13067