def main()

in script/generate_demo.py [0:0]

• 62 lines of code
• 15 McCabe index (conditional complexity)

def main():
    parser = argparse.ArgumentParser()
    parser.add_argument(
        "--input_file",
        default="",
        type=str,
        help="",
    )    
    parser.add_argument(
        "--compared_file",
        default="",
        type=str,
        help="",
    )
    parser.add_argument(
        "--data_file",
        default="",
        type=str,
        help="",
    )
    parser.add_argument(
        "--tasks",
        default="",
        type=str,
        help="",
    )
    parser.add_argument(
        "--num_samples",
        default=100,
        type=int,
        help="",
    )
    args = parser.parse_args()
    with open('vlbert_tasks.yml', 'r') as f:
        task_cfg = edict(yaml.load(f))

    # 1: find the file where input file is correct while target file is wrong
    # 2: get the image id and image path.
    # 3: generate or copy the images. 
    # 4: Show based on different tasks. 
    #   - VQA: image, question, and different answers. 
    #   - RefCOCO: image, caption, and with different bounding box. 
    #   - VCR: image, question, answers. 
    #   - Retrieval: caption, and top 10 retrieved images. 

    input_file = json.load(open(args.input_file, 'r'))
    
    compared_file = None
    if args.compared_file != ''
        compared_file = json.load(open(args.compared_file, 'r'))

    task = 'TASK' + args.tasks
    task_names = task_cfg[task]['name']
    topK = 5
    if task in ['TASK9']:
        with jsonlines.open(args.data_file) as reader:
            entries = []
            imgid2entry = {}
            count = 0
            for annotation in reader:
                image_path = annotation['img_path']
                image_id = int(annotation['img_path'].split('.')[0])
                imgid2entry[image_id] = []
                for sentences in annotation['sentences']:
                    entries.append({"caption": sentences, 'image_id':image_id})
                    imgid2entry[image_id].append(count)
                    count += 1

        # identify which one is correct and which one is wrong.
        for i, seq in enumerate(input_file):
            correct_input = False
            target = int(i / 5)
            if target in seq[:topK]
                correct_input = True

        if compared_file:

        
        pdb.set_trace()