qwenvl2_car_inference.yml

nodes:
  - id: mickrobot_chassis # Run on car
    _unstable_deploy:
      machine: ec7
    # build: wget -nc https://github.com/dora-rs/gosim-2024/releases/download/v0.0.1/robot # cargo install --git https://github.com/lyon-dora/chongyou-car
    path: robot
    inputs:
      text: dora-qwenvl/tick

  - id: camera1 # Run on car
    _unstable_deploy:
      machine: ec7
    # build: pip install opencv-video-capture
    path: opencv-video-capture
    inputs:
      tick: dora/timer/millis/200
    outputs:
      - image
    env:
      CAPTURE_PATH: 0
      IMAGE_WIDTH: 640
      IMAGE_HEIGHT: 480
      ENCODING: jpeg
      FLIP: VERTICAL

  - id: camera2 # Run on car
    _unstable_deploy:
      machine: ec7
    # build: pip install opencv-video-capture
    path: opencv-video-capture
    inputs:
      tick: dora/timer/millis/200
    outputs:
      - image
    env:
      CAPTURE_PATH: 2
      IMAGE_WIDTH: 640
      IMAGE_HEIGHT: 480
      ENCODING: jpeg
      FLIP: VERTICAL

  - id: plot
    # build: |
    # wget -nc https://github.com/dora-rs/gosim-2024/releases/download/v0.0.1/rerun
    # wget -nc https://github.com/dora-rs/gosim-2024/releases/download/v0.0.1/dora-rerun
    path: dora-rerun
    inputs:
      image_1:
        source: camera1/image
        queue_size: 1
      image_2:
        source: camera2/image
        queue_size: 1
      textlog_vlm: dora-qwenvl-recorder/text
    env:
      IMAGE_WIDTH: 640
      IMAGE_HEIGHT: 480
      RERUN_MEMORY_LIMIT: 25%

  - id: dora-qwenvl
    build: pip install dora-qwenvl
    path: dora-qwenvl
    inputs:
      image_1:
        source: camera1/image
        queue_size: 1
      image_2:
        source: camera2/image
        queue_size: 1
      tick: dora/timer/millis/200
    outputs:
      - text
      - tick
    env:
      DEFAULT_QUESTION: Respond with right, left, forward, backward, stop to move to the person
      USE_MODELSCOPE_HUB: True
      ADAPTER_PATH: /openbayes/home/gosim-2024/sft