TEST #755 RESULTS
04/22/2025 - 4:35 AM
ACCELERATOR
MODEL
47.8
tokens/s
generation
1.18
sec
time to first token
1233
tokens/s
prompt
369
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
ARMv8 Processor rev 1 (v8l)
RAM
7.4GB
OS
Linux
Kernel Release
5.15.148-tegra
Architecture
aarch64
Version
Cosmopolitan 3.9.7 MODE=aarch64; #1 SMP PREEMPT Thu Dec 12 15:37:09 PST 2024
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1491
tokens/s
50.0
tokens/s
707
ms
pp4096+tg256
939
tokens/s
41.1
tokens/s
4.39
sec
pp2048+tg256
1254
tokens/s
46.5
tokens/s
1.65
sec
pp2048+tg768
1254
tokens/s
45.8
tokens/s
1.66
sec
pp1024+tg1024
1493
tokens/s
48.5
tokens/s
706
ms
pp1280+tg3072
1404
tokens/s
44.6
tokens/s
932
ms
pp384+tg1152
1731
tokens/s
50.5
tokens/s
241
ms
pp64+tg1024
1452
tokens/s
52.0
tokens/s
63
ms
pp16+tg1536
77
tokens/s
51.1
tokens/s
231
ms