TEST #558 RESULTS
04/08/2025 - 6:23 AM
ACCELERATOR
MODEL
201
tokens/s
generation
89
ms
time to first token
16292
tokens/s
prompt
3323
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 5900X 12-Core Processor (znver3)
RAM
127.9GB
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
21505
tokens/s
197
tokens/s
53
ms
pp4096+tg256
12119
tokens/s
204
tokens/s
343
ms
pp2048+tg256
17697
tokens/s
201
tokens/s
121
ms
pp2048+tg768
17689
tokens/s
201
tokens/s
121
ms
pp1024+tg1024
21949
tokens/s
200
tokens/s
52
ms
pp1280+tg3072
20836
tokens/s
203
tokens/s
66
ms
pp384+tg1152
23962
tokens/s
200
tokens/s
21
ms
pp64+tg1024
8433
tokens/s
200
tokens/s
13
ms
pp16+tg1536
2435
tokens/s
200
tokens/s
12
ms