TEST #148 RESULTS
03/28/2025 - 10:24 PM
ACCELERATOR
MODEL
45.8
tokens/s
generation
772
ms
time to first token
1748
tokens/s
prompt
470
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon CPU E5-2699 v4 @ 2.20GHz (broadwell)
RAM
1007.7GB
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
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2030
tokens/s
47.4
tokens/s
526
ms
pp4096+tg256
1554
tokens/s
41.4
tokens/s
2.66
sec
pp2048+tg256
1851
tokens/s
45.3
tokens/s
1.13
sec
pp2048+tg768
1842
tokens/s
44.7
tokens/s
1.13
sec
pp1024+tg1024
2038
tokens/s
46.4
tokens/s
524
ms
pp1280+tg3072
1978
tokens/s
43.9
tokens/s
669
ms
pp384+tg1152
2306
tokens/s
47.4
tokens/s
187
ms
pp64+tg1024
1730
tokens/s
48.2
tokens/s
57
ms
pp16+tg1536
405
tokens/s
47.7
tokens/s
60
ms