Created by: foreversolar

    def _build_net(self):
        cls_feats = fluid.layers.dropout(
            x=self.feature,
            dropout_prob=0.1,
            dropout_implementation="upscale_in_train")

        if self.hidden_units is not None:
            for n_hidden in self.hidden_units:
                cls_feats = fluid.layers.fc(
                    input=cls_feats, size=n_hidden, act="relu")
        fc0 = fluid.layers.fc(input=cls_feats, size=128 * 3)
        gru_h=fluid.layers.dynamic_gru(input=fc0,size=128,is_reverse=False)
        gru_max=fluid.layers.sequence_pool(input=gru_h,pool_type='max')
        gru_max_tanh=fluid.layers.tanh(gru_max)

1.请问这个self.feature 是预训练模型输出的特征吗？ 2.如何把这个输入接入到 fluid.layers.** 这样的网络层呢？ 比如上述代码想在之后接一个GRU，但是提示需要lod_level=1,而feature和fc0的lod_level都是0 报错如下：

Enforce failed. Expected lods.size() -- 1UL, but received lods.size() :0 !- 1UL:1.
Only support one leve1 sequence now. at [/ paddle/ paddle/ fluid/ operators/ math/ sequence2batch.h: 79]