TEST #1 RESULTS
03/19/2025 - 9:50 PM
ACCELERATOR
MODEL
103
tokens/s
generation
358
ms
time to first token
3552
tokens/s
prompt
1008
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7352 24-Core Processor (znver2)
RAM
125.64GB
OS
Linux
Kernel Release
6.8.0-52-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #53~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Jan 15 19:18:46 UTC 2
RUNTIME
Name
llamafile
Version
0.9.1
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
4250
tokens/s
105
tokens/s
251
ms
pp4096+tg256
3477
tokens/s
92.7
tokens/s
1.19
sec
pp2048+tg256
3977
tokens/s
102
tokens/s
525
ms
pp2048+tg768
3984
tokens/s
101
tokens/s
524
ms
pp1024+tg1024
4248
tokens/s
104
tokens/s
251
ms
pp1280+tg3072
4136
tokens/s
98.5
tokens/s
319
ms
pp384+tg1152
4342
tokens/s
106
tokens/s
98
ms
pp64+tg1024
2921
tokens/s
110
tokens/s
30
ms
pp16+tg1536
636
tokens/s
108
tokens/s
34
ms