TEST #367 RESULTS
04/06/2025 - 1:34 PM
ACCELERATOR
MODEL
122
tokens/s
generation
355
ms
time to first token
4089
tokens/s
prompt
1122
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
12th Gen Intel Core i7-12700H (alderlake)
RAM
31.7GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
4964
tokens/s
127
tokens/s
214
ms
pp4096+tg256
3108
tokens/s
107
tokens/s
1.33
sec
pp2048+tg256
4103
tokens/s
122
tokens/s
508
ms
pp2048+tg768
4083
tokens/s
119
tokens/s
510
ms
pp1024+tg1024
4868
tokens/s
126
tokens/s
218
ms
pp1280+tg3072
4525
tokens/s
114
tokens/s
291
ms
pp384+tg1152
5590
tokens/s
128
tokens/s
76
ms
pp64+tg1024
4359
tokens/s
130
tokens/s
22
ms
pp16+tg1536
1198
tokens/s
129
tokens/s
27
ms