HotpotQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
HotpotQA (HQA) (test)
0.311Exact Match
10
HotpotQA Support Fact Sentences (test)
26.8BLEU-4
10
HotPotQA top 1000 samples (test)
38.6F1
10
HotpotQA (dev)
94.4Top-5
10
HotpotQA fullwiki (eval)
55.16F1 Score
9
HotpotQA (official)
34.3Exact Match (EM)
9
HotpotQA
45Accuracy
9
HotpotQA
44.7Exact Match (EM)
9
HotpotQA
1.4Leakage Error
9
HotPotQA (test)
0.464EM
9
HotPotQA (test)
46.4EM
9
HotpotQA
78.6Answer Accuracy (Ans)
9
HotPotQA (test)
1Abs(0)
9
HotPotQA (test)
20.7Accuracy
9
HotpotQA Out-of-Domain (test)
63.5Accuracy
9
HotpotQA
58.17LM Score
9
HotpotQA
75.63Accuracy
9
HotpotQA
60.62F1 Score
9
HotpotQA Distractor
0.85Normalized Average Latency
9
HotpotQA Distractor
0.78Average Edge Budget
9
HotpotQA
55.9ACCR
9
HotpotQA (full)
69.2EM
9
HotpotQA Full v1.1 (train)
82.2F1 Score
9
HotpotQA
89Exact Match Accuracy
9
HotpotQA 2018 (test)
86.9Recall@4
9