Created by: zt991211

我的相关报错源码如下：

if __name__ == '__main__':
    # 创建模型
    img=[]
    img.append(transform_img(np.array(Image.open('/home/aistudio/work/train-images/V0001_0.png'))))
    img = np.array(img).astype('float32')
    DiscROI_size=600
    with fluid.dygraph.guard():
        model = M_net()
        #启动训练过程
        #train(model)
        model_state_dict, _ = fluid.load_dygraph('/home/aistudio/work/mnet30.pdparams')
        model.load_dict(model_state_dict)
        in_var = fluid.dygraph.to_variable(img)
        out_dygraph, static_layer = fluid.dygraph.TracedLayer.trace(model, inputs=[in_var])
        out_static_graph = static_layer([in_var])
        static_layer.save_inference_model(dirname='/home/aistudio/work/mnet_model')
    place = fluid.CPUPlace()
    exe = fluid.Executor(place)
    program, feed_vars, fetch_vars = fluid.io.load_inference_model('/home/aistudio/work/mnet_model',
    executor=exe)
    fetch, = exe.run(program, feed={feed_vars[0]: in_var},fetch_list=fetch_vars)
    print(fetch.shape)

基本是按照文档源码的格式写的，但现在报错了 出现在fluid.io.load_inference_model('/home/aistudio/work/mnet_model', executor=exe)这一行

报错原因如下： 报错在那个exe那里 ValueError：‘type’ to initialized an Operator can not be None。