Temporal Commonsense Reasoning on MCTACO (test)
79.5F1 ScoreALICE
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ALICEProtocol=Best single model2020.12 | 79.5 | 56.5 | |
| ECONETProtocol=Best single model, Backbone=RoBERTa-Large2020.12 | 76.8 | 54.7 | |
| RoBERTa-Large + ECONETProtocol=Average of 3 seeds2020.12 | 76.3 | 52.8 | |
| RoBERTa-LargeProtocol=Average of 3 seeds2020.12 | 75.5 | 50.4 | |
| RoBERTa-Large + GeneratorProtocol=Average of 3 seeds2020.12 | 75.1 | 50.2 | |
| BERT-LargeProtocol=Average of 3 seeds2020.12 | 70.3 | 43.2 | |
| TacoLMProtocol=Average of 3 seeds2020.12 | 69.3 | 40.5 | |
| BERT-Large + ECONETProtocol=Average of 3 seeds2020.12 | 69.2 | 42.3 |