TEST #673 RESULTS
04/10/2025 - 6:35 PM
ACCELERATOR
MODEL
70.0
tokens/s
generation
369
ms
time to first token
3552
tokens/s
prompt
877
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen Threadripper PRO 5975WX 32-Cores (znver3)
RAM
503.5GB
OS
Linux
Kernel Release
6.8.0-52-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #53-Ubuntu SMP PREEMPT_DYNAMIC Sat Jan 11 00:06:25 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
4406
tokens/s
71.0
tokens/s
248
ms
pp4096+tg256
3358
tokens/s
65.2
tokens/s
1.24
sec
pp2048+tg256
3926
tokens/s
69.9
tokens/s
537
ms
pp2048+tg768
3991
tokens/s
69.1
tokens/s
529
ms
pp1024+tg1024
4289
tokens/s
71.1
tokens/s
254
ms
pp1280+tg3072
3962
tokens/s
67.5
tokens/s
338
ms
pp384+tg1152
4511
tokens/s
71.6
tokens/s
114
ms
pp64+tg1024
2719
tokens/s
72.8
tokens/s
37
ms
pp16+tg1536
809
tokens/s
72.2
tokens/s
33
ms