Дистилляция модели
Related Papers
Cross-lingual Distillation for Text Classification
AbstractCross-lingual text classification(CLTC) is the task of classifying documents written in different languages into the same taxonomy of categories. This paper presents a novel approach to CLTC that builds on …...
Data and Model Distillation as a Solution for Domain-transferable Fact Verification
AbstractWhile neural networks produce state-of-the-art performance in several NLP tasks, they generally depend heavily on lexicalized information, which transfer poorly between domains. We present a combination of two strategies to …...
Being Right for Whose Right Reasons?
AbstractExplainability methods are used to benchmark the extent to which model predictions align with human rationales i.e., are ‘right for the right reasons’. Previous work has failed to acknowledge, however, …...
MiniALBERT: Model Distillation via Parameter-Efficient Recursive Transformers
AbstractPre-trained Language Models (LMs) have become an integral part of Natural Language Processing (NLP) in recent years, due to their superior performance in downstream applications. In spite of this resounding …...