TEST #449 RESULTS
04/07/2025 - 12:34 AM
ACCELERATOR
MODEL
49.1
tokens/s
generation
614
ms
time to first token
2267
tokens/s
prompt
566
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 5 4600G with Radeon Graphics (znver2)
RAM
94.1GB
OS
Linux
Kernel Release
6.12.21-amd64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Debian 6.12.21-1 (2025-03-30)
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2705
tokens/s
50.3
tokens/s
398
ms
pp4096+tg256
1892
tokens/s
45.7
tokens/s
2.19
sec
pp2048+tg256
2301
tokens/s
48.5
tokens/s
911
ms
pp2048+tg768
2386
tokens/s
48.2
tokens/s
879
ms
pp1024+tg1024
2693
tokens/s
49.5
tokens/s
400
ms
pp1280+tg3072
2645
tokens/s
47.6
tokens/s
504
ms
pp384+tg1152
3130
tokens/s
50.4
tokens/s
142
ms
pp64+tg1024
2157
tokens/s
51.2
tokens/s
49
ms
pp16+tg1536
490
tokens/s
50.8
tokens/s
52
ms