使用paddle预测时，在同一个exe下执行多个program但是每次执行都在执行最后一个model
Created by: liujiaming19910220
place = fluid.CPUPlace()
exe = fluid.Executor(place)
exe.run(fluid.default_startup_program())
model_list = list()
for model_filename in model_filename_list:
# [infer_program, feeded_var_names, target_var] = fluid.io.load_inference_model(dirname=save_path, executor=exe)
model = fluid.io.load_inference_model(dirname=model_filename, executor=exe)
model_list.append(model)调用的时候每次都是调用最后一个模型