TEST #40 RESULTS
03/28/2025 - 6:20 PM
ACCELERATOR
MODEL
82.2
tokens/s
generation
623
ms
time to first token
2600
tokens/s
prompt
700
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7352 24-Core Processor (znver2)
RAM
127.9GB
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
2955
tokens/s
78.6
tokens/s
357
ms
pp4096+tg256
1647
tokens/s
48.7
tokens/s
2.51
sec
pp2048+tg256
2366
tokens/s
68.7
tokens/s
880
ms
pp2048+tg768
2371
tokens/s
63.5
tokens/s
878
ms
pp1024+tg1024
3032
tokens/s
85.0
tokens/s
358
ms
pp1280+tg3072
2867
tokens/s
59.9
tokens/s
458
ms
pp384+tg1152
3832
tokens/s
101
tokens/s
108
ms
pp64+tg1024
2870
tokens/s
122
tokens/s
42
ms
pp16+tg1536
1461
tokens/s
112
tokens/s
17
ms