TEST #88 RESULTS
03/28/2025 - 9:23 PM
ACCELERATOR
MODEL
45.9
tokens/s
generation
262
ms
time to first token
5328
tokens/s
prompt
977
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 9354 32-Core Processor (znver4)
RAM
755.6GB
OS
Linux
Kernel Release
6.5.0-45-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #45~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Mon Jul 15 16:40:02 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
7478
tokens/s
48.8
tokens/s
158
ms
pp4096+tg256
4500
tokens/s
39.3
tokens/s
939
ms
pp2048+tg256
6033
tokens/s
44.4
tokens/s
362
ms
pp2048+tg768
6042
tokens/s
43.2
tokens/s
361
ms
pp1024+tg1024
7527
tokens/s
46.2
tokens/s
157
ms
pp1280+tg3072
6792
tokens/s
42.3
tokens/s
210
ms
pp384+tg1152
7025
tokens/s
48.7
tokens/s
74
ms
pp64+tg1024
1956
tokens/s
50.7
tokens/s
52
ms
pp16+tg1536
593
tokens/s
49.4
tokens/s
46
ms