from .audio_process import process_audio_info
from .vision_process import (
    extract_vision_info,
    fetch_image,
    fetch_video,
    process_vision_info,
    smart_resize,
)


def process_mm_info(
    conversations,
    use_audio_in_video,
    return_video_kwargs=False,
    return_video_metadata: bool = False,
    image_patch_size: int = 14,
):
    audios = process_audio_info(conversations, use_audio_in_video)
    vision = process_vision_info(
        conversations,
        return_video_kwargs=return_video_kwargs,
        return_video_metadata=return_video_metadata,
        image_patch_size=image_patch_size,
    )
    return (audios,) + vision