TEST #110 RESULTS
03/28/2025 - 9:50 PM
ACCELERATOR
MODEL
46.9
tokens/s
generation
1.22
sec
time to first token
1057
tokens/s
prompt
343
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon CPU E5-2687W 0 @ 3.10GHz (sandybridge)
RAM
188.8GB
OS
Linux
Kernel Release
5.15.0-105-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #115-Ubuntu SMP Mon Apr 15 09:52:04 UTC 2024
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1256
tokens/s
47.6
tokens/s
836
ms
pp4096+tg256
989
tokens/s
32.1
tokens/s
4.17
sec
pp2048+tg256
1146
tokens/s
43.0
tokens/s
1.81
sec
pp2048+tg768
1145
tokens/s
41.6
tokens/s
1.81
sec
pp1024+tg1024
1254
tokens/s
48.8
tokens/s
835
ms
pp1280+tg3072
1185
tokens/s
39.1
tokens/s
1.10
sec
pp384+tg1152
1278
tokens/s
54.3
tokens/s
317
ms
pp64+tg1024
892
tokens/s
59.1
tokens/s
87
ms
pp16+tg1536
371
tokens/s
56.3
tokens/s
58
ms