| #FQuAD | |
| - No official metric | |
| - We use SQuAD official metric as an inspiration | |
| - Use similar stop words removal and other cleaning of SQuAD | |
| - FQuAD as three annotators response, we use the max F1 score over the three reference as the score |
| #FQuAD | |
| - No official metric | |
| - We use SQuAD official metric as an inspiration | |
| - Use similar stop words removal and other cleaning of SQuAD | |
| - FQuAD as three annotators response, we use the max F1 score over the three reference as the score |