TEST #173 RESULTS
03/30/2025 - 8:12 PM
ACCELERATOR
MODEL
37.0
tokens/s
generation
578
ms
time to first token
2360
tokens/s
prompt
532
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 9800X3D 8-Core Processor
RAM
123.5GB
OS
Linux
Kernel Release
5.15.0-134-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #145-Ubuntu SMP Wed Feb 12 20:08:39 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
3008
tokens/s
43.4
tokens/s
363
ms
pp4096+tg256
2017
tokens/s
26.1
tokens/s
2.07
sec
pp2048+tg256
2567
tokens/s
32.8
tokens/s
831
ms
pp2048+tg768
2559
tokens/s
30.7
tokens/s
834
ms
pp1024+tg1024
3013
tokens/s
38.3
tokens/s
365
ms
pp1280+tg3072
2805
tokens/s
30.7
tokens/s
483
ms
pp384+tg1152
3430
tokens/s
42.1
tokens/s
135
ms
pp64+tg1024
1380
tokens/s
45.3
tokens/s
67
ms
pp16+tg1536
456
tokens/s
43.5
tokens/s
56
ms