sample_info/scripts/compute_informativeness.py [69:90]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    args = parser.parse_args()
    print(args)

    # Load data
    train_data, val_data, test_data, _ = load_data_from_arguments(args, build_loaders=False)
    if args.cache_dataset:
        train_data = CacheDatasetWrapper(train_data)
        val_data = CacheDatasetWrapper(val_data)
        test_data = CacheDatasetWrapper(test_data)

    with open(args.config, 'r') as f:
        architecture_args = json.load(f)

    model_class = getattr(methods, args.model_class)

    model = model_class(input_shape=train_data[0][0].shape,
                        architecture_args=architecture_args,
                        l2_reg_coef=args.l2_reg_coef,
                        device=args.device,
                        seed=args.seed)
    model.eval()
    print("Number of parameters: ", utils.get_num_parameters(model))
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



sample_info/scripts/prepare_informativeness_orders_for_data_summarization.py [69:92]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    args = parser.parse_args()
    print(args)

    # Load data
    train_data, val_data, test_data, _ = load_data_from_arguments(args, build_loaders=False)

    if args.cache_dataset:
        train_data = CacheDatasetWrapper(train_data)
        val_data = CacheDatasetWrapper(val_data)
        test_data = CacheDatasetWrapper(test_data)


    with open(args.config, 'r') as f:
        architecture_args = json.load(f)

    model_class = getattr(methods, args.model_class)

    model = model_class(input_shape=train_data[0][0].shape,
                        architecture_args=architecture_args,
                        l2_reg_coef=args.l2_reg_coef,
                        device=args.device,
                        seed=args.seed)
    model.eval()
    print("Number of parameters: ", utils.get_num_parameters(model))
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -