DROP
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
DROP
92.28DROP Accuracy
138
DROP
92.2F1 Score
96
DROP (test)
96.42F1 Score
76
DROP (dev)
88.1F1 Score
63
DROP
48.74F1 Score
49
DROP
87.5F1 Score
45
DROP
89.62Accuracy
43
DROP
89.27Score
42
DROP
86.2F1 Score
35
DROP (test)
76F1 Score
29
DROP
48.68DROP Score
25
DROP
71.59Exact Match (EM)
25
DROP (test)
90.8Accuracy
23
DROP MRQA out-of-domain evaluation
64.9EM
23
Drop
35.03PSNR
21
DROP
0.673DROP Score
20
DROP
0.4Loss
20
DROP
51.53DROP Score
20
DROP nfl
67.69F1 Score
17
DROP
83.49Performance Score
16
DROP
88.6Accuracy
16
DROP
81.08Accuracy
16
DROP (test)
87.9F1 Score
14
DROP MRQA out-of-domain
0.488EM
14
Drop
45.99PSNR
13