TEST #94 RESULTS
03/28/2025 - 9:27 PM
ACCELERATOR
MODEL
97.8
tokens/s
generation
98
ms
time to first token
13517
tokens/s
prompt
2379
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
INTEL XEON PLATINUM 8558 (sapphirerapids)
RAM
3023.5GB
OS
Linux
Kernel Release
6.8.0-55-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #57-Ubuntu SMP PREEMPT_DYNAMIC Wed Feb 12 23:42:21 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
20051
tokens/s
101
tokens/s
61
ms
pp4096+tg256
15717
tokens/s
88.6
tokens/s
276
ms
pp2048+tg256
18076
tokens/s
95.1
tokens/s
124
ms
pp2048+tg768
18165
tokens/s
94.3
tokens/s
124
ms
pp1024+tg1024
17267
tokens/s
95.2
tokens/s
73
ms
pp1280+tg3072
16388
tokens/s
93.6
tokens/s
90
ms
pp384+tg1152
13584
tokens/s
101
tokens/s
42
ms
pp64+tg1024
1898
tokens/s
107
tokens/s
48
ms
pp16+tg1536
510
tokens/s
104
tokens/s
45
ms