Saved in:
Bibliographic Details
Main Authors: Yufan, Zhu, Zeyu, Hao, Siqi, Li, Boqian, Niu
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.05499
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915052841336832
author Yufan, Zhu
Zeyu, Hao
Siqi, Li
Boqian, Niu
author_facet Yufan, Zhu
Zeyu, Hao
Siqi, Li
Boqian, Niu
contents SplaXBERT, built on ALBERT-xlarge with context-splitting and mixed precision training, achieves high efficiency in question-answering tasks on lengthy texts. Tested on SQuAD v1.1, it attains an Exact Match of 85.95% and an F1 Score of 92.97%, outperforming traditional BERT-based models in both accuracy and resource efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2412_05499
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SplaXBERT: Leveraging Mixed Precision Training and Context Splitting for Question Answering
Yufan, Zhu
Zeyu, Hao
Siqi, Li
Boqian, Niu
Computation and Language
Machine Learning
SplaXBERT, built on ALBERT-xlarge with context-splitting and mixed precision training, achieves high efficiency in question-answering tasks on lengthy texts. Tested on SQuAD v1.1, it attains an Exact Match of 85.95% and an F1 Score of 92.97%, outperforming traditional BERT-based models in both accuracy and resource efficiency.
title SplaXBERT: Leveraging Mixed Precision Training and Context Splitting for Question Answering
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2412.05499