TEST #492 RESULTS
04/07/2025 - 11:56 AM
ACCELERATOR
MODEL
51.2
tokens/s
generation
1.44
sec
time to first token
888
tokens/s
prompt
316
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5800X3D 8-Core Processor (znver3)
RAM
31.9GB
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
1082
tokens/s
60.2
tokens/s
963
ms
pp4096+tg256
835
tokens/s
34.0
tokens/s
4.93
sec
pp2048+tg256
983
tokens/s
43.2
tokens/s
2.11
sec
pp2048+tg768
984
tokens/s
41.5
tokens/s
2.11
sec
pp1024+tg1024
1068
tokens/s
50.9
tokens/s
977
ms
pp1280+tg3072
1021
tokens/s
38.0
tokens/s
1.27
sec
pp384+tg1152
1128
tokens/s
60.6
tokens/s
355
ms
pp64+tg1024
384
tokens/s
68.4
tokens/s
178
ms
pp16+tg1536
507
tokens/s
63.8
tokens/s
43
ms