PrismML's Ternary Bonsai 2 27B fits a 24 GB card or Mac, but decodes at ~30 tok/s on an L4 and ~21 on an M4 Pro. z-lab's DFlash 2 drafter was trained on bf16 Qwen3.8-27B, so it guesses worse on the ternary model. I fine-tuned it on 1.5M tokens of Bonsai 2's own greedy output. Drafter (safetensors +
