2 ms·

I've run it on a Pi 5 with 8Gb, and get about a token a second

• theshrike79 2y ago
  M-series are a LOT faster :) Even my M1/16GB gets decent speeds. 7+ tokens/second with llama3