TEST #128 RESULTS
03/28/2025 - 10:10 PM
ACCELERATOR
MODEL
63.4
tokens/s
generation
595
ms
time to first token
2184
tokens/s
prompt
615
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon CPU E5-2650 v4 @ 2.20GHz (broadwell)
RAM
62.6GB
OS
Linux
Kernel Release
6.8.0-45-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #45~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Sep 11 15:25:05 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2593
tokens/s
65.2
tokens/s
410
ms
pp4096+tg256
2062
tokens/s
60.0
tokens/s
2.00
sec
pp2048+tg256
2395
tokens/s
62.2
tokens/s
871
ms
pp2048+tg768
2394
tokens/s
62.5
tokens/s
871
ms
pp1024+tg1024
2593
tokens/s
63.6
tokens/s
410
ms
pp1280+tg3072
2492
tokens/s
60.8
tokens/s
529
ms
pp384+tg1152
2663
tokens/s
65.5
tokens/s
159
ms
pp64+tg1024
1974
tokens/s
65.5
tokens/s
48
ms
pp16+tg1536
489
tokens/s
65.6
tokens/s
48
ms