Gemma 4 5.1B E2B
on M4 Pro 64GB · dasLLAMA
Summary
User reports Gemma 4 E2B at 86.6 t/s generation and 1545.7 t/s prompt processing on an M4 Pro with 64GB unified memory.
Setup is dasLLAMA with Q8_0 weights, f16 KV cache, and one Metal GPU; the session requested 640 positions and no speculative decoding.
The figures are native PP512 and TG128 counts over five repetitions, not one 512-in/128-out request.