Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Fact Verification on FEVER (Accuracy and Latency)
Loading...
61.4
Accuracy
ToT
45.176
49.388
53.6
57.812
Jun 13, 2024
Accuracy
Latency (s)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Latency (s)
ToT
Backbone=Mistral-7B
2024.06
61.4
3,291.3
TS-SFT
Backbone=Mistral-7B
2024.06
60.2
41.7
CPO
Backbone=Mistral-7B
2024.06
59.9
40.6
CoT
Backbone=Mistral-7B
2024.06
57.9
41.8
CPO
Backbone=LLAMA2-7B
2024.06
53.2
36.8
CPO
Backbone=LLAMA2-13B
2024.06
49.2
43.9
ToT
Backbone=LLAMA2-13B
2024.06
48.8
3,139.8
TS-SFT
Backbone=LLAMA2-13B
2024.06
48.8
44
CoT
Backbone=LLAMA2-13B
2024.06
48.2
45.4
TS-SFT
Backbone=LLAMA2-7B
2024.06
47.5
34
ToT
Backbone=LLAMA2-7B
2024.06
47.3
2,087.1
CoT
Backbone=LLAMA2-7B
2024.06
45.8
33.8
Feedback
Search any
task
Search any
task