TEST #22 RESULTS
03/27/2025 - 9:40 PM
ACCELERATOR
MODEL
248
tokens/s
generation
135
ms
time to first token
8984
tokens/s
prompt
2545
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon CPU E5-2690 v4 @ 2.60GHz (broadwell)
RAM
688.5GB
OS
Linux
Kernel Release
5.15.0-119-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #129-Ubuntu SMP Fri Aug 2 19:25:20 UTC 2024
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
11266
tokens/s
266
tokens/s
95
ms
pp4096+tg256
9096
tokens/s
194
tokens/s
456
ms
pp2048+tg256
11207
tokens/s
236
tokens/s
187
ms
pp2048+tg768
11355
tokens/s
228
tokens/s
185
ms
pp1024+tg1024
12831
tokens/s
255
tokens/s
84
ms
pp1280+tg3072
11641
tokens/s
218
tokens/s
114
ms
pp384+tg1152
10385
tokens/s
271
tokens/s
41
ms
pp64+tg1024
2266
tokens/s
286
tokens/s
32
ms
pp16+tg1536
810
tokens/s
280
tokens/s
23
ms