List your company
ResearchFormerly WizWand

Combinatory Categorial Grammar on CCG (test)

0.961AccuracyCVT + Multi-task + Large
0.946440.950220.9540.95778Sep 22, 2018
Updated 8mo ago

Evaluation results

MethodLinks
CVT + Multi-task + LargeSemi-supervised=true, Multi-task=true, Model Size=Large2018.090.961
CVT + Multi-Task + LargeSupervision Type=Semi-supervised, Learning Paradigm=Multi-task, Model Scale=Large2018.090.9605
CVT + Multi-taskSemi-supervised=true, Multi-task=true2018.090.96
CVT + Multi-TaskSupervision Type=Semi-supervised, Learning Paradigm=Multi-task, Model Scale=Standard2018.090.9597
ELMo + Multi-taskSupervision Type=Semi-supervised, Learning Paradigm=Multi-task, Model Scale=Standard2018.090.9591
ELMo + Multi-taskSemi-supervised=true, Multi-task=true2018.090.959
ELMo (our implementation)Semi-supervised=true, Multi-task=false2018.090.958
ELMoSupervision Type=Semi-supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.090.9579
CVTSemi-supervised=true, Multi-task=false2018.090.957
CVTSupervision Type=Semi-supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.090.9565
Word DropoutSemi-supervised=true, Multi-task=false2018.090.952
Word DropoutSupervision Type=Semi-supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.090.952
Shortcut LSTMSemi-supervised=false, Multi-task=false2018.090.951
Virtual Adversarial TrainingSemi-supervised=true, Multi-task=false2018.090.951
Shortcut LSTMSupervision Type=Supervised2018.090.9508
Virtual Adversarial TrainingSupervision Type=Semi-supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.090.9507
SupervisedSupervision Type=Supervised, Learning Paradigm=Single-task, Model Scale=Standard2018.090.9494
SupervisedSemi-supervised=false, Multi-task=false2018.090.949
Tri-Trained LSTMSupervision Type=Semi-supervised2018.090.947