List your company
ResearchFormerly WizWand

Part-of-Speech Tagging on POS (test)

97.79AccuracyELMo + Multi-task
90.364492.292294.2296.1478Sep 22, 2018Jul 2, 2019Apr 10, 2020Jan 18, 2021Oct 28, 2021Aug 7, 2022May 18, 2023
Updated 8mo ago

Evaluation results

MethodLinks
ELMo + Multi-taskSemi-supervised=true, Multi-task=true2018.0997.79
ELMo + Multi-taskSupervision Type=Semi-supervised, Learning Paradigm=Multi-task, Model Scale=Standard2018.0997.79
CVT + Multi-taskSemi-supervised=true, Multi-task=true2018.0997.76
CVT + Multi-TaskSupervision Type=Semi-supervised, Learning Paradigm=Multi-task, Model Scale=Standard2018.0997.76
CVT + Multi-task + LargeSemi-supervised=true, Multi-task=true, Model Size=Large2018.0997.74
CVT + Multi-Task + LargeSupervision Type=Semi-supervised, Learning Paradigm=Multi-task, Model Scale=Large2018.0997.74
ELMo (our implementation)Semi-supervised=true, Multi-task=false2018.0997.72
ELMoSupervision Type=Semi-supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.0997.72
CVTSemi-supervised=true, Multi-task=false2018.0997.7
CVTSupervision Type=Semi-supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.0997.7
Word DropoutSemi-supervised=true, Multi-task=false2018.0997.66
Word DropoutSupervision Type=Semi-supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.0997.66
Virtual Adversarial TrainingSemi-supervised=true, Multi-task=false2018.0997.64
Virtual Adversarial TrainingSupervision Type=Semi-supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.0997.64
SupervisedSemi-supervised=false, Multi-task=false2018.0997.6
SupervisedSupervision Type=Supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.0997.6
JMTSemi-supervised=false, Multi-task=true2018.0997.55
LSTM-CNN-CRFSupervision Type=Supervised2018.0997.55
JMT*Supervision Type=Semi-supervised, Learning Paradigm=Multi-task2018.0997.55
Shortcut LSTMSemi-supervised=false, Multi-task=false2018.0997.53
Shortcut LSTMSupervision Type=Supervised2018.0997.53
LM-LSTM-CNN-CRFSupervision Type=Supervised2018.0997.53
ZEN (P)Initialization=Pre-trained, Backbone Scale=Base2019.1196.64
BERT (P)Initialization=Pre-trained, Backbone Scale=Base2019.1196.1
ZEN (R)Initialization=Random, Backbone Scale=Base2019.1195.82
BERT (R)Initialization=Random, Backbone Scale=Base2019.1195.43
Ref.2023.0592.28
TAPIR-Trfdelay=22023.0591.49
TAPIR-Trfdelay=12023.0591.35
TAPIR-Trfdelay=02023.0591.32
TAPIR-LTdelay=02023.0590.9
TAPIR-LTdelay=12023.0590.83
TAPIR-LTdelay=22023.0590.65