Created by: JayEworld

网络结构代码 user_vec = paddle.layer.fc(name='user_vec', input = user_fc3, size = label_dict['out_sz'], act = paddle.activation.Linear()) user_vec = paddle.layer.batch_norm(name='user_vec_bn', input = user_vec, act = paddle.activation.Relu())

用capi预测的时候， paddle_gradient_machine_get_layer_output(this->_predict_machines[m_idx], "user_vec", out_args);

请问输出的是bn之前的输出还是bn之后的输出呢？？？