Temporal Machine Reading Comprehension on TORQUE (test)
76.3F1 ScoreECONET
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ECONETProtocol=Best single model, Backbone=RoBERTa-Large2020.12 | 76.3 | 52 | 37 | |
| RoBERTa-Large + ECONETProtocol=Average of 3 seeds2020.12 | 76.1 | 51.6 | 36.8 | |
| RoBERTa-Large + GeneratorProtocol=Average of 3 seeds2020.12 | 75.8 | 51.2 | 35.8 | |
| RoBERTa-Large (Ning et al.)Protocol=Average of 3 seeds2020.12 | 75.2 | 51.1 | 34.5 | |
| RoBERTa-LargeProtocol=Average of 3 seeds2020.12 | 75.1 | 49.6 | 35.3 | |
| BERT-Large + ECONETProtocol=Average of 3 seeds2020.12 | 71.4 | 44.8 | 28.5 | |
| BERT-LargeProtocol=Average of 3 seeds2020.12 | 70.6 | 43.7 | 27.5 | |
| TacoLMProtocol=Average of 3 seeds2020.12 | 65.4 | 37.1 | 21 |