HyperAI超神经

首页资讯最新论文教程数据集百科 SOTA LLM 模型天梯 GPU 天梯顶会

中文

HyperAI超神经

Question Answering On Quora Question Pairs

评估指标

Accuracy

评测结果

各个模型在此基准测试上的表现结果

模型名称	Accuracy	Paper Title	Repository
T5-11B	90.4%	Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
24hBERT	70.7	How to Train BERT with an Academic Budget
MLM+ subs+ del-span	90.3%	CLEAR: Contrastive Learning for Sentence Representation	-
ELECTRA	90.1%	ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators
RoBERTa (ensemble)	90.2%	RoBERTa: A Robustly Optimized BERT Pretraining Approach
T5-Small	88.0%	Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
ERNIE 2.0 Large	90.1%	ERNIE 2.0: A Continual Pre-training Framework for Language Understanding
BigBird	88.6%	Big Bird: Transformers for Longer Sequences
T5-Base	89.4%	Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
RE2	89.2 %	Simple and Effective Text Matching with Richer Alignment Features
SqueezeBERT	80.3%	SqueezeBERT: What can computer vision teach NLP about efficient neural networks?
DeBERTa (large)	92.3%	DeBERTa: Decoding-enhanced BERT with Disentangled Attention
ALBERT	90.5%	ALBERT: A Lite BERT for Self-supervised Learning of Language Representations
XLNet (single model)	92.3%	XLNet: Generalized Autoregressive Pretraining for Language Understanding
SWEM-concat	83.03%	Baseline Needs More Love: On Simple Word-Embedding-Based Models and Associated Pooling Mechanisms
T5-3B	89.7%	Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
T5-Large 770M	89.9%	Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
DistilBERT 66M	89.2%	DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter
ERNIE 2.0 Base	89.8%	ERNIE 2.0: A Continual Pre-training Framework for Language Understanding

0 of 19 row(s) selected.