Created by: Superjomn

After saving the inference model, it seems no way to get the output of internal operators.

[inference_program, feed_target_names, fetch_targets] = \
        fluid.io.load_inference_model(save_dirname, exe)
results = exe.run(inference_program,
                              feed={feed_target_name: tensor_x},
                              fetch_list=fetch_targets)