List your company
ResearchFormerly WizWand

Text Classification on Yahoo! Answer 500 labels

0.3092Top-1 Error RateSoftMatch/FreeMatch + SemiReward
0.3064880.3247940.34310.361406May 22, 2023Jun 13, 2023Jul 6, 2023Jul 28, 2023Aug 20, 2023Sep 11, 2023Oct 4, 2023
Updated 8mo ago

Evaluation results

MethodLinks
SoftMatch/FreeMatch + SemiRewardBackbone=Bert [Devlin et al., 2018], SemiReward=true2023.100.30920.99
AdaMatch# Labels=500, Backbone=Bert-Base2023.050.3275—
FreeMatch# Labels=500, Backbone=Bert-Base2023.050.3277—
SoftMatch/FreeMatchBackbone=Bert [Devlin et al., 2018], SemiReward=false2023.100.3302—
FixMatch# Labels=500, Backbone=Bert-Base2023.050.3303—
SimMatch# Labels=500, Backbone=Bert-Base2023.050.3306—
SoftMatch# Labels=500, Backbone=Bert-Base2023.050.3307—
CoMatch# Labels=500, Backbone=Bert-Base2023.050.3348—
FlexMatch + SemiRewardBackbone=Bert [Devlin et al., 2018], SemiReward=true2023.100.33640.99
ILL# Labels=500, Backbone=Bert-Base2023.050.338—
FlexMatchBackbone=Bert [Devlin et al., 2018], SemiReward=false2023.100.3473—
Pseudo LabelBackbone=Bert [Devlin et al., 2018], SemiReward=false2023.100.3487—
VAT# Labels=500, Backbone=Bert-Base2023.050.3487—
Pseudo Label + SemiRewardBackbone=Bert [Devlin et al., 2018], SemiReward=true2023.100.35080.99
Dash# Labels=500, Backbone=Bert-Base2023.050.3526—
FlexMatch# Labels=500, Backbone=Bert-Base2023.050.3561—
MixMatch# Labels=500, Backbone=Bert-Base2023.050.3575—
Mean-Teacher# Labels=500, Backbone=Bert-Base2023.050.3709—
Pseudo-Label# Labels=500, Backbone=Bert-Base2023.050.377—