TEST #108 RESULTS
03/28/2025 - 9:47 PM
ACCELERATOR
MODEL
160
tokens/s
generation
301
ms
time to first token
4298
tokens/s
prompt
1317
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon CPU E5-1650 v3 @ 3.50GHz (haswell)
RAM
31.2GB
OS
Linux
Kernel Release
5.15.0-131-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #141-Ubuntu SMP Fri Jan 10 21:18:28 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
5307
tokens/s
150
tokens/s
200
ms
pp4096+tg256
3973
tokens/s
109
tokens/s
1.04
sec
pp2048+tg256
4751
tokens/s
148
tokens/s
438
ms
pp2048+tg768
4728
tokens/s
144
tokens/s
440
ms
pp1024+tg1024
5231
tokens/s
169
tokens/s
201
ms
pp1280+tg3072
4835
tokens/s
134
tokens/s
271
ms
pp384+tg1152
5535
tokens/s
187
tokens/s
74
ms
pp64+tg1024
3128
tokens/s
202
tokens/s
25
ms
pp16+tg1536
1188
tokens/s
194
tokens/s
18
ms