TEST #519 RESULTS
04/07/2025 - 8:03 PM
ACCELERATOR
MODEL
90.1
tokens/s
generation
274
ms
time to first token
4789
tokens/s
prompt
1164
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i5-14600K (alderlake)
RAM
63.8GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
6108
tokens/s
91.7
tokens/s
179
ms
pp4096+tg256
4333
tokens/s
83.8
tokens/s
958
ms
pp2048+tg256
5356
tokens/s
89.4
tokens/s
394
ms
pp2048+tg768
5383
tokens/s
88.5
tokens/s
392
ms
pp1024+tg1024
6015
tokens/s
91.1
tokens/s
181
ms
pp1280+tg3072
5935
tokens/s
87.4
tokens/s
227
ms
pp384+tg1152
6416
tokens/s
92.5
tokens/s
71
ms
pp64+tg1024
2570
tokens/s
93.8
tokens/s
35
ms
pp16+tg1536
985
tokens/s
92.8
tokens/s
27
ms