TEST #70 RESULTS
03/28/2025 - 9:01 PM
ACCELERATOR
MODEL
42.9
tokens/s
generation
669
ms
time to first token
2056
tokens/s
prompt
509
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7713 64-Core Processor (znver3)
RAM
503.5GB
OS
Linux
Kernel Release
6.5.0-44-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #44~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Tue Jun 18 14:36:16 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2394
tokens/s
44.3
tokens/s
451
ms
pp4096+tg256
1762
tokens/s
39.7
tokens/s
2.35
sec
pp2048+tg256
2161
tokens/s
42.6
tokens/s
972
ms
pp2048+tg768
2152
tokens/s
42.1
tokens/s
976
ms
pp1024+tg1024
2505
tokens/s
43.4
tokens/s
432
ms
pp1280+tg3072
2379
tokens/s
41.4
tokens/s
561
ms
pp384+tg1152
2749
tokens/s
43.8
tokens/s
163
ms
pp64+tg1024
1986
tokens/s
44.4
tokens/s
54
ms
pp16+tg1536
421
tokens/s
44.1
tokens/s
59
ms