deploy/docker-compose.yaml

services:
  vllm-iquest-optimized:
    image: vllm-blackwell-official:latest
    container_name: vllm-iquest-nvfp4-hq
    environment:
      - VLLM_USE_V1=1
      - NVIDIA_VISIBLE_DEVICES=all
      - VLLM_LOGGING_LEVEL=DEBUG
    volumes:
      - ../NVFP4-Packed:/model
    ports:
      - "8001:8000"
    command: >
      --model /model --served-model-name iquest-coder-40b-loop --quantization modelopt --trust-remote-code --tensor-parallel-size 1 --gpu-memory-utilization 0.8 --max-model-len 32768 --enforce-eager
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [ gpu ]

  benchmarking:
    image: vllm-blackwell-official:latest
    container_name: iquest-benchmark
    volumes:
      - ../NVFP4-Packed:/model
      - .:/scripts
    entrypoint: python3
    command: /scripts/benchmark_optimized.py
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [ gpu ]