Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Natural Language Generation on Math
Loading...
38.8
BLEU
ADORE
29.96
32.255
34.55
36.845
Jul 2, 2024
BLEU
ROUGE
BERT-F
Updated 3d ago
Evaluation Results
Method
Method
Links
BLEU
ROUGE
BERT-F
ADORE
Backbone=Llama-2 7B, C...
2024.07
38.8
28.9
70.5
Full Attention
Backbone=Llama-2 7B, C...
2024.07
38.6
29.9
69.7
H2O(Rebuilt)
Backbone=Llama-2 7B, C...
2024.07
34.7
27.1
69.6
H2O
Backbone=Llama-2 7B, C...
2024.07
33.3
26.2
68.1
Strided Attention
Backbone=Llama-2 7B, C...
2024.07
33
26.7
66.7
StreamingLLM
Backbone=Llama-2 7B, C...
2024.07
32.9
26.8
68.3
KV Compression
Backbone=Llama-2 7B, C...
2024.07
32.2
24.4
66.4
Window Attention
Backbone=Llama-2 7B, C...
2024.07
30.3
24.3
66.3
Feedback
Search any
task
Search any
task