الرئيسية الأخبار أحدث الأوراق البحثية الدروس مجموعات البيانات الموسوعة SOTA نماذج LLM لوحة الأداء GPU الفعاليات

العربية

Relation Classification On Tacred 1

المقاييس

F1

النتائج

نتائج أداء النماذج المختلفة على هذا المعيار القياسي

اسم النموذج	F1	Paper Title	Repository
BERT	66.0	ERNIE: Enhanced Language Representation with Informative Entities	-
TANL (multi-task)	61.9	Structured Prediction as Translation between Augmented Natural Languages	-
LUKE 483M	72.7	LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention	-
DeepEx (zero-shot top-10)	76.4	Zero-Shot Information Extraction as a Unified Text-to-Triple Translation	-
KnowBERT	71.5	Knowledge Enhanced Contextual Word Representations	-
DeepEx (zero-shot top-1)	49.2	Zero-Shot Information Extraction as a Unified Text-to-Triple Translation	-
MTB Baldini Soares et al. (2019)	71.5	Matching the Blanks: Distributional Similarity for Relation Learning	-
Deepstruct zero-shot	36.1	DeepStruct: Pretraining of Language Models for Structure Prediction	-
SpanBERT	70.8	SpanBERT: Improving Pre-training by Representing and Predicting Spans	-
RoBERTa	71.3	K-Adapter: Infusing Knowledge into Pre-Trained Models with Adapters	-
DeepStruct multi-task w/ finetune	76.8	DeepStruct: Pretraining of Language Models for Structure Prediction	-
ERNIE	68.0	ERNIE: Enhanced Language Representation with Informative Entities	-
DeepStruct multi-task	74.9	DeepStruct: Pretraining of Language Models for Structure Prediction	-
TANL	71.9	Structured Prediction as Translation between Augmented Natural Languages	-
KEPLER	71.7	KEPLER: A Unified Model for Knowledge Embedding and Pre-trained Language Representation	-
C-GCN	66.4	Graph Convolution over Pruned Dependency Trees Improves Relation Extraction	-
K-Adapter	72.0	K-Adapter: Infusing Knowledge into Pre-Trained Models with Adapters	-

0 of 17 row(s) selected.

Relation Classification On Tacred 1 | SOTA | HyperAI