MODEL=Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-Q4_K_P.gguf
DRAFT=Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-FastMTP-32K.gguf
DEPTH=3

./llama-server \
  --model "$MODEL" \
  --spec-draft-model "$DRAFT" \
  --spec-draft-ngl all \
  --spec-type draft-mtp \
  --spec-draft-n-max "$DEPTH" \
  --spec-draft-p-min 0 \
  --ctx-size 131072 \
  --parallel 1 \
  --batch-size 2048 \
  --ubatch-size 512 \
  --n-gpu-layers all \
  --split-mode none \
  --flash-attn on \
  --no-mmap \
  --temp 0.7 \
  --top-k 20 \
  --top-p 0.95 \
  --min-p 0 \
  --presence-penalty 0 \
  --repeat-penalty 1.0 \
  --jinja \
  --reasoning on \
  --reasoning-effort low \
  --reasoning-budget 256 \
  --no-reasoning-preserve \
  --reasoning-format deepseek \
  --host 127.0.0.1 \
  --port 8081 \
  --mmproj mmproj-Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-BF16.gguf