#!/bin/bash
python -m vllm.entrypoints.openai.api_server \
  --model qwen/Qwen3-32B-Instruct \
  --max-model-len 65536 \
  --kv-cache-dtype fp8 \
  --rope-scaling '{"rope_type":"yarn","factor":2.0,"original_max_position_embeddings":32768}' \
  --max-num-seqs 8 \
  --max-num-batched-tokens 131072 \
  --gpu-memory-utilization 0.90 \
  --speculative-config '{"method":"ngram","num_speculative_tokens":5}' \
  --host 0.0.0.0 \
  --port 8000