Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Model Compression on Llama 8B 3.1
Loading...
2.8
Model Size (GB)
UniQL
2.272
5.836
9.4
12.964
Dec 3, 2025
Model Size (GB)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Model Size (GB)
UniQL
W-bit=4, Prun. p%=35%
2025.12
2.8
UniQL
W-bit=4, Prun. p%=0%
2025.12
4.1
TAO-HQQ
W-bit=4*, Prun. p%=0%
2025.12
5.7
TRT-AWQ
W-bit=4*, Prun. p%=0%
2025.12
5.8
MoDeGPT
W-bit=16, Prun. p%=15%
2025.12
13.9
SVD-LLM
W-bit=16, Prun. p%=15%
2025.12
14.1
FP16
W-bit=16, Prun. p%=0%
2025.12
16
Feedback
Search any
task
Search any
task
