TEST #64 RESULTS
03/28/2025 - 8:55 PM
ACCELERATOR
MODEL
81.9
tokens/s
generation
117
ms
time to first token
10806
tokens/s
prompt
1964
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 9374F 32-Core Processor (znver4)
RAM
1511.6GB
OS
Linux
Kernel Release
6.8.0-40-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #40~22.04.3-Ubuntu SMP PREEMPT_DYNAMIC Tue Jul 30 17:30:19 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
14325
tokens/s
87.0
tokens/s
83
ms
pp4096+tg256
12478
tokens/s
74.4
tokens/s
342
ms
pp2048+tg256
13642
tokens/s
80.2
tokens/s
163
ms
pp2048+tg768
13649
tokens/s
79.4
tokens/s
163
ms
pp1024+tg1024
14236
tokens/s
81.9
tokens/s
84
ms
pp1280+tg3072
12751
tokens/s
78.0
tokens/s
113
ms
pp384+tg1152
12345
tokens/s
84.9
tokens/s
43
ms
pp64+tg1024
2699
tokens/s
86.0
tokens/s
36
ms
pp16+tg1536
1127
tokens/s
85.1
tokens/s
25
ms