Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
LLM Decoding on Llama-2-70B
Loading...
0.2163
Per-step Decoding Latency
Pre3
0.211712
0.242681
0.27365
0.304619
Jun 4, 2025
Per-step Decoding Latency
Updated 4d ago
Evaluation Results
Method
Method
Links
Per-step Decoding Latency
Pre3
Batchsize=1
2025.06
0.2163
Pre3
Batchsize=4
2025.06
0.2407
XGrammar
Batchsize=1
2025.06
0.303
XGrammar
Batchsize=4
2025.06
0.331
Feedback
Search any
task
Search any
task
