TEST #967 RESULTS
06/01/2025 - 8:44 PM
ACCELERATOR
MODEL
103
tokens/s
generation
489
ms
time to first token
3192
tokens/s
prompt
875
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5800X 8-Core Processor (znver3)
RAM
15.9GB
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
3854
tokens/s
122
tokens/s
275
ms
pp4096+tg256
2092
tokens/s
62.0
tokens/s
1.97
sec
pp2048+tg256
3012
tokens/s
85.5
tokens/s
692
ms
pp2048+tg768
3007
tokens/s
79.8
tokens/s
693
ms
pp1024+tg1024
3843
tokens/s
101
tokens/s
275
ms
pp1280+tg3072
3620
tokens/s
70.9
tokens/s
363
ms
pp384+tg1152
4559
tokens/s
122
tokens/s
91
ms
pp64+tg1024
3101
tokens/s
147
tokens/s
26
ms
pp16+tg1536
1637
tokens/s
135
tokens/s
15
ms