Text Classification on Hyperpartisan (val)
87.8Precision (%)Fine-Tuning
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Fine-Tuning% Tuned=100, Backbone=Longformer-base2023.05 | 87.8 | 76.2 | 81.5 | |
| Prefix-Propagation% Tuned=0.05, Backbone=Longformer-base2023.05 | 86.4 | 77.7 | 81.8 | |
| Prefix-Tuning% Tuned=0.1, Backbone=Longformer-base2023.05 | 78.3 | 73.8 | 75.3 | |
| ROBERTa PT% Tuned=0.1, Backbone=ROBERTa-base2023.05 | 70.4 | 59.2 | 64.1 |