TEST #490 RESULTS
04/07/2025 - 11:22 AM
ACCELERATOR
MODEL
36.4
tokens/s
generation
694
ms
time to first token
1821
tokens/s
prompt
457
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 9950X3D 16-Core Processor
RAM
60.4GB
OS
Linux
Kernel Release
6.13.8-zen1-1-zen
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 ZEN SMP PREEMPT_DYNAMIC Sun, 23 Mar 2025 17:17:17 +0000
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2376
tokens/s
40.7
tokens/s
456
ms
pp4096+tg256
1763
tokens/s
28.5
tokens/s
2.36
sec
pp2048+tg256
2105
tokens/s
34.8
tokens/s
1.00
sec
pp2048+tg768
2112
tokens/s
33.6
tokens/s
998
ms
pp1024+tg1024
2392
tokens/s
37.1
tokens/s
453
ms
pp1280+tg3072
2240
tokens/s
32.3
tokens/s
598
ms
pp384+tg1152
2356
tokens/s
39.4
tokens/s
189
ms
pp64+tg1024
579
tokens/s
41.1
tokens/s
133
ms
pp16+tg1536
468
tokens/s
40.1
tokens/s
58
ms