TEST #427 RESULTS
04/06/2025 - 9:23 PM
ACCELERATOR
MODEL
38.3
tokens/s
generation
874
ms
time to first token
1600
tokens/s
prompt
413
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q8_0
SYSTEM
CPU
AMD Ryzen 9 5950X 16-Core Processor (znver3)
RAM
125.7GB
OS
Linux
Kernel Release
6.13.9-200.fc41.x86_64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Sat Mar 29 01:29:31 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1964
tokens/s
42.7
tokens/s
545
ms
pp4096+tg256
1294
tokens/s
24.5
tokens/s
3.21
sec
pp2048+tg256
1661
tokens/s
37.4
tokens/s
1.26
sec
pp2048+tg768
1653
tokens/s
35.8
tokens/s
1.27
sec
pp1024+tg1024
1900
tokens/s
40.3
tokens/s
563
ms
pp1280+tg3072
1890
tokens/s
32.0
tokens/s
701
ms
pp384+tg1152
2120
tokens/s
43.2
tokens/s
203
ms
pp64+tg1024
1377
tokens/s
45.2
tokens/s
68
ms
pp16+tg1536
545
tokens/s
43.9
tokens/s
51
ms