TEST #770 RESULTS
04/26/2025 - 4:06 AM
ACCELERATOR
MODEL
20.1
tokens/s
generation
4.66
sec
time to first token
306
tokens/s
prompt
110
LocalScore
HOW YOU STACK UP
Explore All Resultsqwen2.5-3b-instruct - Q4_0
SYSTEM
CPU
Intel Core i7-10750H CPU @ 2.60GHz (skylake)
RAM
23.3GB
OS
Linux
Kernel Release
6.13.12-200.fc41.x86_64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Sun Apr 20 15:52:43 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
328
tokens/s
24.8
tokens/s
3.16
sec
pp4096+tg256
250
tokens/s
8.1
tokens/s
16.54
sec
pp2048+tg256
299
tokens/s
14.3
tokens/s
6.91
sec
pp2048+tg768
297
tokens/s
13.1
tokens/s
6.97
sec
pp1024+tg1024
338
tokens/s
19.3
tokens/s
3.07
sec
pp1280+tg3072
330
tokens/s
11.5
tokens/s
3.92
sec
pp384+tg1152
362
tokens/s
27.2
tokens/s
1.08
sec
pp64+tg1024
348
tokens/s
34.9
tokens/s
199
ms
pp16+tg1536
204
tokens/s
27.6
tokens/s
93
ms