Question Answering on KILT: HotpotQA

Metric: EM (higher is better)

LeaderboardDataset

Loading chart...

Results

Sort:

#	Model↕	EM▼	Extra Data	Paper	Date↕	Code
1	intersect	40.46	No	-	-	-
2	Wikipedia	36.9	No	-	-	-
3	Multitask DPR + BART	31.77	No	-	-	-
4	Sphere	31.64	No	-	-	-
5	RAG	26.97	No	-	-	-
6	BART + DPR	25.18	No	-	-	-
7	BART	15.37	No	-	-	-
8	T5-base	12.64	No	KILT: a Benchmark for Knowledge Intensive Langua...	2020-09-04	Code
9	BERT + DPR	11.29	No	-	-	-
10	multi-task small	3.29	No	-	-	-
11	TABi	0	No	-	-	-
12	chriskuei	0	No	-	-	-
13	GENRE	0	No	-	-	-
14	Multi-task DPR	0	No	-	-	-

#1intersect
40.46
EM
No paper
#2Wikipedia
36.9
EM
No paper
#3Multitask DPR + BART
31.77
EM
No paper
#4Sphere
31.64
EM
No paper
#5RAG
26.97
EM
No paper
#6BART + DPR
25.18
EM
No paper
#7BART
15.37
EM
No paper
#8T5-baseSOTA
12.64
EM· 2020-09-04
KILT: a Benchmark for Knowledge Intensive Language Tasks Code
#9BERT + DPR
11.29
EM
No paper
#10multi-task small
3.29
EM
No paper
#11TABi
0
EM
No paper
#12chriskuei
0
EM
No paper
#13GENRE
0
EM
No paper
#14Multi-task DPR
0
EM
No paper