TEST #59 RESULTS
03/28/2025 - 8:49 PM
ACCELERATOR
MODEL
43.8
tokens/s
generation
979
ms
time to first token
1330
tokens/s
prompt
390
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7352 24-Core Processor (znver2)
RAM
251.6GB
OS
Linux
Kernel Release
6.8.0-45-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #45~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Sep 11 15:25:05 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1540
tokens/s
45.2
tokens/s
688
ms
pp4096+tg256
1279
tokens/s
39.3
tokens/s
3.23
sec
pp2048+tg256
1435
tokens/s
43.1
tokens/s
1.45
sec
pp2048+tg768
1431
tokens/s
42.5
tokens/s
1.45
sec
pp1024+tg1024
1542
tokens/s
44.5
tokens/s
686
ms
pp1280+tg3072
1494
tokens/s
41.7
tokens/s
879
ms
pp384+tg1152
1559
tokens/s
45.6
tokens/s
268
ms
pp64+tg1024
1304
tokens/s
46.4
tokens/s
92
ms
pp16+tg1536
384
tokens/s
46.0
tokens/s
62
ms