Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Quantization on LLAMA v1 (train)
Loading...
0.25
Processing Time (hr)
OPTQ
-1.8876
12.5412
26.97
41.3988
Feb 14, 2024
Processing Time (hr)
Updated 4d ago
Evaluation Results
Method
Method
Links
Processing Time (hr)
OPTQ
Target=layer-wise reco...
2024.02
0.25
OPTQ
Target=layer-wise reco...
2024.02
0.45
OPTQ
Target=layer-wise reco...
2024.02
1.08
Z-FOLD
Target=layer-wise reco...
2024.02
1.13
OmniQuant
Target=attention-wise...
2024.02
2.37
Z-FOLD
Target=layer-wise reco...
2024.02
2.48
OmniQuant
Target=attention-wise...
2024.02
4.2
aespa
Target=attention-wise...
2024.02
6.84
OmniQuant
Target=attention-wise...
2024.02
9.84
Z-FOLD
Target=layer-wise reco...
2024.02
10.51
aespa
Target=attention-wise...
2024.02
15.89
AffineQuant
Target=attention-wise...
2024.02
18.76
AffineQuant
Target=attention-wise...
2024.02
18.76
AffineQuant
Target=attention-wise...
2024.02
47.84
aespa
Target=attention-wise...
2024.02
53.69
Feedback
Search any
task
Search any
task