TEST #663 RESULTS
04/10/2025 - 2:17 AM
ACCELERATOR
MODEL
95.6
tokens/s
generation
225
ms
time to first token
6191
tokens/s
prompt
1380
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen AI 9 HX 370 w/ Radeon 890M
RAM
30.5GB
OS
Linux
Kernel Release
6.14.1-cachyos1.fc41.x86_64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Wed Apr 9 01:53:25 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
7999
tokens/s
108
tokens/s
138
ms
pp4096+tg256
5181
tokens/s
71.8
tokens/s
806
ms
pp2048+tg256
6550
tokens/s
90.6
tokens/s
325
ms
pp2048+tg768
6530
tokens/s
88.6
tokens/s
326
ms
pp1024+tg1024
7668
tokens/s
97.2
tokens/s
144
ms
pp1280+tg3072
7048
tokens/s
83.2
tokens/s
193
ms
pp384+tg1152
9572
tokens/s
104
tokens/s
52
ms
pp64+tg1024
4025
tokens/s
110
tokens/s
24
ms
pp16+tg1536
1150
tokens/s
107
tokens/s
22
ms