Apple M2 Max (38c, 96GB) llamafile: Llama-3.1-8B Q4_K_M prompt-eval
484.06 tok/s prompt-eval (batch); decode 46.41 t/s
Fields
id
benchmarks/localscore-m2-max-38c-96gb-llama31-8b-q4km-ppMetric
prefill_tok_s
Value
484.06
Unit
tok/s prompt-eval (batch); decode 46.41 t/s
Model
llama-3-1-8b-instruct Q4_K_M
Methodology
localscore.ai accelerator/515 measured row via siliconscore CSV.
Measured by
third_party