TEST #51 RESULTS
03/28/2025 - 8:43 PM
ACCELERATOR
MODEL
285
tokens/s
generation
131
ms
time to first token
10328
tokens/s
prompt
2820
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
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
12727
tokens/s
298
tokens/s
84
ms
pp4096+tg256
8760
tokens/s
246
tokens/s
472
ms
pp2048+tg256
11183
tokens/s
280
tokens/s
187
ms
pp2048+tg768
11139
tokens/s
275
tokens/s
188
ms
pp1024+tg1024
13202
tokens/s
292
tokens/s
81
ms
pp1280+tg3072
12205
tokens/s
266
tokens/s
108
ms
pp384+tg1152
14783
tokens/s
300
tokens/s
29
ms
pp64+tg1024
8041
tokens/s
306
tokens/s
11
ms
pp16+tg1536
911
tokens/s
302
tokens/s
21
ms