TEST #3 RESULTS
03/19/2025 - 9:54 PM
ACCELERATOR
MODEL
330
tokens/s
generation
95
ms
time to first token
14061
tokens/s
prompt
3662
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
17261
tokens/s
325
tokens/s
62
ms
pp4096+tg256
12325
tokens/s
285
tokens/s
336
ms
pp2048+tg256
15731
tokens/s
327
tokens/s
133
ms
pp2048+tg768
15825
tokens/s
321
tokens/s
133
ms
pp1024+tg1024
18598
tokens/s
337
tokens/s
58
ms
pp1280+tg3072
17052
tokens/s
311
tokens/s
78
ms
pp384+tg1152
19412
tokens/s
348
tokens/s
23
ms
pp64+tg1024
9284
tokens/s
361
tokens/s
10
ms
pp16+tg1536
1061
tokens/s
355
tokens/s
18
ms