TEST #411 RESULTS
04/06/2025 - 7:18 PM
ACCELERATOR
MODEL
173
tokens/s
generation
185
ms
time to first token
7408
tokens/s
prompt
1907
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5800X 8-Core Processor (znver3)
RAM
31.9GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
9273
tokens/s
177
tokens/s
116
ms
pp4096+tg256
6117
tokens/s
165
tokens/s
676
ms
pp2048+tg256
7925
tokens/s
175
tokens/s
264
ms
pp2048+tg768
7979
tokens/s
176
tokens/s
262
ms
pp1024+tg1024
9326
tokens/s
176
tokens/s
116
ms
pp1280+tg3072
8841
tokens/s
172
tokens/s
151
ms
pp384+tg1152
10447
tokens/s
173
tokens/s
42
ms
pp64+tg1024
5488
tokens/s
172
tokens/s
17
ms
pp16+tg1536
1271
tokens/s
173
tokens/s
18
ms