Text Classification on AG News 40 labels
0.1067Top-1 Error RateSoftMatch/FreeMatch + SemiReward
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SoftMatch/FreeMatch + SemiRewardBackbone=Bert [Devlin et al., 2018], SemiReward=true2023.10 | 0.1067 | 0.93 | |
| SoftMatch/FreeMatchBackbone=Bert [Devlin et al., 2018], SemiReward=false2023.10 | 0.1169 | — | |
| AdaMatch# Labels=40, Backbone=Bert-Base2023.05 | 0.1173 | — | |
| SoftMatch# Labels=40, Backbone=Bert-Base2023.05 | 0.119 | — | |
| CoMatch# Labels=40, Backbone=Bert-Base2023.05 | 0.1195 | — | |
| FlexMatch + SemiRewardBackbone=Bert [Devlin et al., 2018], SemiReward=true2023.10 | 0.126 | 0.93 | |
| FreeMatch# Labels=40, Backbone=Bert-Base2023.05 | 0.1298 | — | |
| FlexMatchBackbone=Bert [Devlin et al., 2018], SemiReward=false2023.10 | 0.1308 | — | |
| MixMatch# Labels=40, Backbone=Bert-Base2023.05 | 0.135 | — | |
| Pseudo Label + SemiRewardBackbone=Bert [Devlin et al., 2018], SemiReward=true2023.10 | 0.139 | 0.93 | |
| SimMatch# Labels=40, Backbone=Bert-Base2023.05 | 0.1426 | — | |
| VAT# Labels=40, Backbone=Bert-Base2023.05 | 0.147 | — | |
| ILL# Labels=40, Backbone=Bert-Base2023.05 | 0.1477 | — | |
| Mean-Teacher# Labels=40, Backbone=Bert-Base2023.05 | 0.1517 | — | |
| Pseudo LabelBackbone=Bert [Devlin et al., 2018], SemiReward=false2023.10 | 0.1519 | — | |
| FlexMatch# Labels=40, Backbone=Bert-Base2023.05 | 0.1638 | — | |
| Dash# Labels=40, Backbone=Bert-Base2023.05 | 0.1767 | — | |
| Pseudo-Label# Labels=40, Backbone=Bert-Base2023.05 | 0.1949 | — | |
| FixMatch# Labels=40, Backbone=Bert-Base2023.05 | 0.3017 | — |