Created by: moranliuyu

• 版本、环境信息：    1）PaddlePaddle版本：1.5.0    2）CPU：    3）GPU：    4）系统环境：windows
• 模型信息

class ActorModel(Model):
    def __init__(self, act_dim):
        hid1_size = 400
        hid2_size = 300

        self.fc1 = layers.fc(size=hid1_size, act='relu')
        self.fc2 = layers.fc(size=hid2_size, act='relu')
        self.fc3 = layers.fc(size=act_dim, act='tanh')

    def policy(self, obs):
        hid1 = self.fc1(obs)
        hid2 = self.fc2(hid1)
        means = self.fc3(hid2)
        means = layers.argmax(means, axis=1)
        means = layers.unsqueeze(means, axes=[1])
        return means.astype('float32')

cost = layers.reduce_mean(-1.0 * Q)
optimizer = fluid.optimizer.AdamOptimizer(self.actor_lr)
 optimizer.minimize(cost, parameter_list=self.model.get_actor_params())

• 问题描述： 如上定义好模型（用了argmax()）后，然后定义目标函数和优化函数，在进入到 optimizer.minimize(cost, parameter_list=self.model.get_actor_params())时系统报错： Process finished with exit code -1073741819 (0xC0000005)

请问：对于argmax()梯度反向传播的问题，是不是不能这么定义网络？