TEST #408 RESULTS
04/06/2025 - 7:09 PM
ACCELERATOR
MODEL
169
tokens/s
generation
191
ms
time to first token
7163
tokens/s
prompt
1851
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5800X 8-Core Processor (znver3)
RAM
31.9GB
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
9069
tokens/s
171
tokens/s
119
ms
pp4096+tg256
5940
tokens/s
161
tokens/s
696
ms
pp2048+tg256
7780
tokens/s
165
tokens/s
269
ms
pp2048+tg768
7432
tokens/s
171
tokens/s
282
ms
pp1024+tg1024
9128
tokens/s
173
tokens/s
118
ms
pp1280+tg3072
8564
tokens/s
170
tokens/s
155
ms
pp384+tg1152
10142
tokens/s
172
tokens/s
44
ms
pp64+tg1024
5169
tokens/s
170
tokens/s
19
ms
pp16+tg1536
1247
tokens/s
171
tokens/s
19
ms