[docheinestages]

Any estimates of the performance (prompt processing and decoding tokens/s) on consumer hardware like Macbook Pro M-series?