How to run inference with mutli GPU?

[AIisCool]

def parse_args():
    parser = ArgumentParser()
    parser.add_argument(
        '--img', type=str, default=None, help='Input image file.')
    parser.add_argument(
        '--video', type=str, default=None, help='Input video file.')
    parser.add_argument(
        '--label',
        type=int,
        default=None,
        help='Input label for conditional models.')
    parser.add_argument(
        '--trimap', type=str, default=None, help='Input for matting models.')
    parser.add_argument(
        '--mask',
        type=str,
        default=None,
        help='path to input mask file for inpainting models')
    parser.add_argument(
        '--text',
        type=str,
        default='',
        help='text input for text2image models')
    parser.add_argument(
        '--result-out-dir',
        type=str,
        default=None,
        help='Output img or video path.')
    parser.add_argument(
        '--model-name',
        type=str,
        default=None,
        help='Pretrained mmagic algorithm')
    parser.add_argument(
        '--model-setting',
        type=int,
        default=None,
        help='Pretrained mmagic algorithm setting')
    parser.add_argument(
        '--model-config',
        type=str,
        default=None,
        help='Path to the custom config file of the selected mmagic model.')
    parser.add_argument(
        '--model-ckpt',
        type=str,
        default=None,
        help='Path to the custom checkpoint file of the selected det model.')
    parser.add_argument(
        '--device',
        type=str,
        default='cuda',
        help='Device used for inference.')
    parser.add_argument(
        '--extra-parameters',
        nargs='+',
        action=DictAction,
        help='Other customized kwargs for different model')
    parser.add_argument(
        '--seed',
        type=int,
        default=2022,
        help='The random seed used in inference.')

There is no argument to have mutli GPU?

How can I made this use mutli GPU?

python demo/mmagic_inference_demo.py \
        --model-name basicvsr_pp \
        --model-config configs/basicvsr_pp/basicvsr-pp_c128n25_600k_ntire-decompress-track3.py \
        --model-ckpt C:/mmagic-main/models/basicvsr_pp/basicvsr_plusplus_c128n25_ntire_decompress_track3_20210304-6daf4a40.pth \
        --device cuda \
        --video C:/mmagic-main/data/INPUT/video.mp4 \
        --result-out-dir C:/mmagic-main/data/OUTPUT/videoframes

Thank you🙏