TEST #7 RESULTS
03/19/2025 - 10:16 PM
ACCELERATOR
MODEL
218
tokens/s
generation
161
ms
time to first token
9337
tokens/s
prompt
2328
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7352 24-Core Processor (znver2)
RAM
125.64GB
OS
Linux
Kernel Release
6.8.0-52-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #53~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Jan 15 19:18:46 UTC 2
RUNTIME
Name
llamafile
Version
0.9.1
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
11458
tokens/s
230
tokens/s
94
ms
pp4096+tg256
6469
tokens/s
199
tokens/s
638
ms
pp2048+tg256
9310
tokens/s
217
tokens/s
225
ms
pp2048+tg768
9243
tokens/s
212
tokens/s
226
ms
pp1024+tg1024
11732
tokens/s
221
tokens/s
92
ms
pp1280+tg3072
10780
tokens/s
205
tokens/s
123
ms
pp384+tg1152
15581
tokens/s
225
tokens/s
29
ms
pp64+tg1024
7280
tokens/s
228
tokens/s
13
ms
pp16+tg1536
2181
tokens/s
225
tokens/s
12
ms