[Docathon][Delete Deprecated API Doc No.10-11] (#7269)

* Delete docs/api/paddle/static/nn/multi_box_head_cn.rst 删除废弃API multi_box_head_cn.rst * Delete docs/api/paddle/static/nn/sequence_concat_cn.rst 删除废弃API sequence_concat_cn.rst * Update Overview_cn.rst 删除废弃API对应的文档

Author: Alex-Yanggg

docs/api/paddle/static/Overview_cn.rst

@@ -68,13 +68,11 @@ Executor 相关 API
  " :ref:`group_norm <cn_api_paddle_static_nn_group_norm>` ", "Group Normalization 方法"
  " :ref:`instance_norm <cn_api_paddle_static_nn_instance_norm>` ", "Instance Normalization 方法"
  " :ref:`layer_norm <cn_api_paddle_static_nn_layer_norm>` ", "Layer Normalization 方法"
- " :ref:`multi_box_head <cn_api_paddle_static_nn_multi_box_head>` ", "SSD 检测头 "
  " :ref:`nce <cn_api_paddle_static_nn_nce>` ", "计算并返回噪音对比估计损失"
  " :ref:`prelu <cn_api_paddle_static_nn_prelu>` ", "prelu 激活函数"
  " :ref:`row_conv <cn_api_paddle_static_nn_row_conv>` ", "行卷积"
  " :ref:`spectral_norm <cn_api_paddle_static_nn_spectral_norm>` ", "Spectral Normalization 方法"
  " :ref:`switch_case <cn_api_paddle_static_nn_switch_case>` ", "类似于 c++的 switch/case"
- " :ref:`sequence_concat <cn_api_paddle_static_nn_sequence_concat>` ", "仅支持带有 LoD 信息的 Tensor ，通过 Tensor 的 LoD 信息将输入的多个 Tensor 进行连接，输出连接后的 Tensor"
  " :ref:`sequence_conv <cn_api_paddle_static_nn_sequence_conv>` ", "仅支持带有 LoD 信息的 Tensor，在给定的卷积参数下，对输入的变长序列 Tensor 进行卷积操作"
  " :ref:`sequence_expand <cn_api_paddle_static_nn_sequence_expand>` ", "仅支持带有 LoD 信息的 Tensor，根据输入 y 的第 ref_level 层 lod 对输入 x 进行扩展"
  " :ref:`sequence_first_step <cn_api_paddle_static_nn_sequence_first_step>` ", "仅支持带有 LoD 信息的 Tensor，对输入的 Tensor，在最后一层 lod_level 上，选取其每个序列的第一个时间步的特征向量作为池化后的输出向量"