Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Block-wise quantization complexity analysis on OPT Models
Loading...
0.24
GFLOPS
aespa
-1.3904
9.6148
20.62
31.6252
Feb 14, 2024
GFLOPS
Updated 4d ago
Evaluation Results
Method
Method
Links
GFLOPS
aespa
Model size=125M
2024.02
0.24
aespa
Model size=350M
2024.02
0.42
aespa
Model size=1.3B
2024.02
1.6
aespa
Model size=2.7B
2024.02
3.2
Conventional block-wise quantization
Model size=125M, Batch...
2024.02
6.7
Conventional block-wise quantization
Model size=350M, Batch...
2024.02
7.5
Conventional block-wise quantization
Model size=1.3B, Batch...
2024.02
11
aespa
Model size=6.7B
2024.02
13
Conventional block-wise quantization
Model size=2.7B, Batch...
2024.02
15
aespa
Model size=13B
2024.02
20
Conventional block-wise quantization
Model size=6.7B, Batch...
2024.02
34
Conventional block-wise quantization
Model size=13B, Batch...
2024.02
41
Feedback
Search any
task
Search any
task