TEST #444 RESULTS
04/06/2025 - 11:45 PM
ACCELERATOR
MODEL
68.0
tokens/s
generation
350
ms
time to first token
3916
tokens/s
prompt
913
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
13th Gen Intel Core i7-13700K (alderlake)
RAM
63.7GB
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
4785
tokens/s
68.2
tokens/s
229
ms
pp4096+tg256
3338
tokens/s
65.2
tokens/s
1.24
sec
pp2048+tg256
4220
tokens/s
68.4
tokens/s
500
ms
pp2048+tg768
4146
tokens/s
67.7
tokens/s
509
ms
pp1024+tg1024
4762
tokens/s
68.8
tokens/s
230
ms
pp1280+tg3072
4754
tokens/s
66.6
tokens/s
283
ms
pp384+tg1152
5474
tokens/s
69.1
tokens/s
84
ms
pp64+tg1024
2903
tokens/s
68.8
tokens/s
36
ms
pp16+tg1536
860
tokens/s
68.9
tokens/s
33
ms