TEST #355 RESULTS
04/06/2025 - 5:53 AM
ACCELERATOR
MODEL
99.0
tokens/s
generation
84
ms
time to first token
15261
tokens/s
prompt
2617
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 9554 64-Core Processor (znver4)
RAM
1511.5GB
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
20240
tokens/s
102
tokens/s
60
ms
pp4096+tg256
17674
tokens/s
90.9
tokens/s
244
ms
pp2048+tg256
19357
tokens/s
97.2
tokens/s
116
ms
pp2048+tg768
19406
tokens/s
96.5
tokens/s
116
ms
pp1024+tg1024
20377
tokens/s
97.6
tokens/s
60
ms
pp1280+tg3072
17884
tokens/s
94.2
tokens/s
82
ms
pp384+tg1152
17366
tokens/s
101
tokens/s
31
ms
pp64+tg1024
4029
tokens/s
107
tokens/s
24
ms
pp16+tg1536
1017
tokens/s
104
tokens/s
26
ms