TEST #220 RESULTS
04/04/2025 - 12:35 AM
ACCELERATOR
MODEL
55.7
tokens/s
generation
352
ms
time to first token
3834
tokens/s
prompt
847
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 3700X 8-Core Processor (znver2)
RAM
63.9GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
4897
tokens/s
49.4
tokens/s
235
ms
pp4096+tg256
3383
tokens/s
58.0
tokens/s
1.23
sec
pp2048+tg256
4104
tokens/s
55.1
tokens/s
519
ms
pp2048+tg768
4376
tokens/s
57.4
tokens/s
486
ms
pp1024+tg1024
4921
tokens/s
56.1
tokens/s
225
ms
pp1280+tg3072
4812
tokens/s
55.6
tokens/s
283
ms
pp384+tg1152
5353
tokens/s
56.9
tokens/s
88
ms
pp64+tg1024
2149
tokens/s
56.8
tokens/s
53
ms
pp16+tg1536
511
tokens/s
56.4
tokens/s
48
ms