Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Conversational Question Answering on CANARD
Loading...
0.681
F1 Score
EXCORD
0.56972
0.59861
0.6275
0.65639
Jun 22, 2021
F1 Score
HEQ-Q Score
HEQ-D Deviation
Updated 1mo ago
Evaluation Results
Method
Method
Links
F1 Score
HEQ-Q Score
HEQ-D Deviation
EXCORD
Backbone=RoBERTa
2021.06
0.681
0.642
8.4
RoBERTa
Approach=Pipeline
2021.06
0.669
0.632
7.3
RoBERTa
Approach=End-to-end
2021.06
0.658
0.622
7.1
EXCORD
Backbone=BERT+HAE
2021.06
0.631
0.584
5.7
EXCORD
Backbone=BERT
2021.06
0.626
0.582
6.4
BERT+HAE
Approach=Pipeline
2021.06
0.624
0.578
6
BERT
Approach=Pipeline
2021.06
0.622
0.578
6
BERT+HAE
Approach=End-to-end
2021.06
0.582
0.535
5.5
BERT
Approach=End-to-end
2021.06
0.574
0.529
3.2
Feedback
Search any
task
Search any
task
