Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Saligram, Pradyumna, Lanpouthakoun, Andrew
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2407.14039
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916330254368768
author Saligram, Pradyumna
Lanpouthakoun, Andrew
author_facet Saligram, Pradyumna
Lanpouthakoun, Andrew
contents We explore advanced fine-tuning techniques to boost BERT's performance in sentiment analysis, paraphrase detection, and semantic textual similarity. Our approach leverages SMART regularization to combat overfitting, improves hyperparameter choices, employs a cross-embedding Siamese architecture for improved sentence embeddings, and introduces innovative early exiting methods. Our fine-tuning findings currently reveal substantial improvements in model efficiency and effectiveness when combining multiple fine-tuning architectures, achieving a state-of-the-art performance score of on the test set, surpassing current benchmarks and highlighting BERT's adaptability in multifaceted linguistic tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2407_14039
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle BERTer: The Efficient One
Saligram, Pradyumna
Lanpouthakoun, Andrew
Computation and Language
Machine Learning
We explore advanced fine-tuning techniques to boost BERT's performance in sentiment analysis, paraphrase detection, and semantic textual similarity. Our approach leverages SMART regularization to combat overfitting, improves hyperparameter choices, employs a cross-embedding Siamese architecture for improved sentence embeddings, and introduces innovative early exiting methods. Our fine-tuning findings currently reveal substantial improvements in model efficiency and effectiveness when combining multiple fine-tuning architectures, achieving a state-of-the-art performance score of on the test set, surpassing current benchmarks and highlighting BERT's adaptability in multifaceted linguistic tasks.
title BERTer: The Efficient One
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2407.14039