TEST #850 RESULTS
05/10/2025 - 10:07 AM
MODEL
4.2
tokens/s
generation
128.28
sec
time to first token
12
tokens/s
prompt
7
LocalScore
HOW YOU STACK UP
Explore All ResultsDeepSeek R1 Distill Qwen 1.5B - Q4_K - Medium
SYSTEM
CPU
RAM
15.6GB
OS
Linux
Kernel Release
6.1.43-rockchip-rk3588
Architecture
aarch64
Version
Cosmopolitan 3.9.7 MODE=aarch64; #1.1.8 SMP Fri Feb 2 21:16:10 CST 2024
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
11
tokens/s
4.2
tokens/s
90.46
sec
pp4096+tg256
9
tokens/s
3.7
tokens/s
438.23
sec
pp2048+tg256
11
tokens/s
4.0
tokens/s
192.50
sec
pp2048+tg768
11
tokens/s
3.8
tokens/s
193.05
sec
pp1024+tg1024
11
tokens/s
3.8
tokens/s
96.76
sec
pp1280+tg3072
12
tokens/s
4.2
tokens/s
107.94
sec
pp384+tg1152
13
tokens/s
4.5
tokens/s
29.29
sec
pp64+tg1024
14
tokens/s
4.6
tokens/s
4.76
sec
pp16+tg1536
12
tokens/s
4.5
tokens/s
1.52
sec