TEST #18 RESULTS
03/25/2025 - 5:55 PM
ACCELERATOR
MODEL
83.9
tokens/s
generation
617
ms
time to first token
2606
tokens/s
prompt
708
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7352 24-Core Processor (znver2)
RAM
127.87GB
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
3054
tokens/s
96.3
tokens/s
346
ms
pp4096+tg256
1667
tokens/s
48.6
tokens/s
2.48
sec
pp2048+tg256
2373
tokens/s
68.1
tokens/s
877
ms
pp2048+tg768
2378
tokens/s
66.5
tokens/s
875
ms
pp1024+tg1024
3019
tokens/s
82.4
tokens/s
362
ms
pp1280+tg3072
2870
tokens/s
59.8
tokens/s
457
ms
pp384+tg1152
3830
tokens/s
102
tokens/s
108
ms
pp64+tg1024
2828
tokens/s
120
tokens/s
29
ms
pp16+tg1536
1434
tokens/s
111
tokens/s
19
ms