TEST #27 RESULTS
03/27/2025 - 11:46 PM
ACCELERATOR
MODEL
86.6
tokens/s
generation
457
ms
time to first token
2817
tokens/s
prompt
811
LocalScore
HOW YOU STACK UP
Explore All ResultsMinistral 8B Instruct 2410 TEST - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7H12 64-Core Processor (znver2)
RAM
503.5GB
OS
Linux
Kernel Release
6.8.0-52-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #53~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Jan 15 19:18:46 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
3367
tokens/s
90.6
tokens/s
315
ms
pp4096+tg256
2712
tokens/s
77.2
tokens/s
1.52
sec
pp2048+tg256
3093
tokens/s
85.2
tokens/s
674
ms
pp2048+tg768
3086
tokens/s
83.8
tokens/s
676
ms
pp1024+tg1024
3310
tokens/s
88.1
tokens/s
321
ms
pp1280+tg3072
3218
tokens/s
82.0
tokens/s
409
ms
pp384+tg1152
3350
tokens/s
89.8
tokens/s
126
ms
pp64+tg1024
2594
tokens/s
91.7
tokens/s
35
ms
pp16+tg1536
621
tokens/s
90.9
tokens/s
36
ms