SQuAD
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
SQuAD (test)
79.87Exact Match (EM)
6
SQuAD 2.0
91.71Score
6
SQuAD
0.837Accuracy
6
SQUAD 1.1 (Memorization)
59F1 Score
6
SQuAD v2
86.3Accuracy
6
SQuAD 2K
84.9Answer F1 Score
6
SQuAD 1K
85.1Answer F1
6
SQuAD 512
0.859Answer F1
6
SQuAD-Shifts New Wiki
5.18MAPE
6
SQUAD
74Exact Match (EM)
6
SQuAD QG
24.59BLEU-4
6
SQuAD MRQA
92.7F1 Score
6
SQuAD (val)
89.4F1
6
SQuAD Double-Char (test)
69.52EM
5
SQuAD Keyboard-Char (test)
69.04EM
5
SQuAD v1.1 (val)
41.6F1 Score
5
SQuAD 2.0
58.8AUROC
5
SQuAD 100 passages (val)
35.4Immediate Accuracy
5
SQuAD sampled from original dataset
1.291Average Length (s)
5
SQuAD
0.78E-SC
5
SQuAD
6.7E-SC
5
SQuAD v2 (test held-out)
0.858AUC
5
SQuAD 1.1, 2.0 (dev)
76SQuAD 1.1 EM/F1 Score
5
SQuAD (test)
89.2F1
5
SQuAD
35.93EM
5