Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Latency Profiling on Mamba2-8B Nano 8G (inference)
Loading...
55.3
TPOT (Time Per Output Token)
UniQL
54.248
61.349
68.45
75.551
Dec 3, 2025
TPOT (Time Per Output Token)
TTLT (Total Latency)
Updated 1mo ago
Evaluation Results
Method
Method
Links
TPOT (Time Per Output Token)
TTLT (Total Latency)
UniQL
W-bit=4, Prun. p%=35%
2025.12
55.3
28,508.1
UniQL
W-bit=4, Prun. p%=0%
2025.12
81.6
41,116.3
Feedback
Search any
task
Search any
task
