TEST #376 RESULTS
04/06/2025 - 2:18 PM
ACCELERATOR
MODEL
4.3
tokens/s
generation
8.73
sec
time to first token
290
tokens/s
prompt
52
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
12th Gen Intel Core i7-12700H (alderlake)
RAM
31.7GB
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
146
tokens/s
5.3
tokens/s
7.20
sec
pp4096+tg256
77
tokens/s
2.2
tokens/s
53.57
sec
pp2048+tg256
466
tokens/s
5.0
tokens/s
4.60
sec
pp2048+tg768
445
tokens/s
4.8
tokens/s
4.81
sec
pp1024+tg1024
453
tokens/s
4.8
tokens/s
2.47
sec
pp1280+tg3072
388
tokens/s
4.1
tokens/s
3.55
sec
pp384+tg1152
386
tokens/s
4.2
tokens/s
1.23
sec
pp64+tg1024
202
tokens/s
4.2
tokens/s
549
ms
pp16+tg1536
47
tokens/s
4.2
tokens/s
575
ms