clean dirty index

demo-serving/data/model/paddle/fluid/ctr_prediction/__model__.prototxt

+blocks {
+  idx: 0
+  parent_idx: -1
+  vars {
+    name: "sequence_pool_17.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C18"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_20.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_16.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C6"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_16.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_21.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_15.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_11.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_14.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_11.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_14.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C12"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C2"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_12.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_1.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_11.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C21"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_10.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_1.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_15.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_1.b_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: 400
+        }
+      }
+    }
+    persistable: true
+  }
+  vars {
+    name: "sequence_pool_1.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C8"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_0.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_3.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 2
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_3.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 2
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_3.b_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: 2
+        }
+      }
+    }
+    persistable: true
+  }
+  vars {
+    name: "fc_2.w_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: 400
+          dims: 400
+        }
+      }
+    }
+    persistable: true
+  }
+  vars {
+    name: "fc_2.tmp_2"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 400
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C5"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_2.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 400
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_1.tmp_2"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 400
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_16.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_3.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_15.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_13.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C13"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_14.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_10.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_10.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C25"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_13.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fetch"
+    type {
+      type: FETCH_LIST
+    }
+    persistable: true
+  }
+  vars {
+    name: "embedding_6.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_19.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "dense_input"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 13
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_24.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_3.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_7.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_18.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_7.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_4.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_9.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_18.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C20"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C17"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_6.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_7.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_24.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_24.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_4.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_6.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_2.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_18.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C19"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_5.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_8.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_20.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_0.w_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: 273
+          dims: 400
+        }
+      }
+    }
+    persistable: true
+  }
+  vars {
+    name: "C7"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_22.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_9.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_19.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_21.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_4.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C9"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_5.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_9.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_0.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_17.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_22.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C23"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C16"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_2.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 400
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_12.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_25.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_21.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_20.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_2.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C24"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_17.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "feed"
+    type {
+      type: FEED_MINIBATCH
+    }
+    persistable: true
+  }
+  vars {
+    name: "sequence_pool_13.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_3.tmp_2"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 2
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C3"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C14"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_3.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C10"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_3.w_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: 400
+          dims: 2
+        }
+      }
+    }
+    persistable: true
+  }
+  vars {
+    name: "embedding_23.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_22.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_0.tmp_2"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 400
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C4"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_0.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C22"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_23.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C11"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "save_infer_model/scale_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 2
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_19.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C15"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_23.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_25.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_8.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_0.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 400
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "C26"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: INT64
+          dims: -1
+          dims: 1
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_5.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_0.b_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: 400
+        }
+      }
+    }
+    persistable: true
+  }
+  vars {
+    name: "sequence_pool_25.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "embedding_2.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 10
+        }
+        lod_level: 1
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_0.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 400
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "concat_0.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 273
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_1.tmp_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 400
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "sequence_pool_8.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_1.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: -1
+          dims: 400
+        }
+        lod_level: 0
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_1.w_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: 400
+          dims: 400
+        }
+      }
+    }
+    persistable: true
+  }
+  vars {
+    name: "sequence_pool_12.tmp_1"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+        }
+      }
+    }
+    persistable: false
+  }
+  vars {
+    name: "fc_2.b_0"
+    type {
+      type: LOD_TENSOR
+      lod_tensor {
+        tensor {
+          data_type: FP32
+          dims: 400
+        }
+      }
+    }
+    persistable: true
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "dense_input"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 26
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_25.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 25
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_24.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 24
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_23.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 23
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_22.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 22
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_21.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 21
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_20.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 20
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_19.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 19
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_18.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 18
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_17.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 17
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_16.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 16
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_15.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 15
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_14.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 14
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_13.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 13
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_12.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 12
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_11.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 11
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_10.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 10
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_9.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 9
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_8.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 8
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_7.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 7
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_6.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 6
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_5.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 5
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_4.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 4
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_3.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 3
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_2.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 2
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_1.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 1
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "feed"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "embedding_0.tmp_0"
+    }
+    type: "feed"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1725, in _prepend_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 845, in prepend_feed_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1000, in save_inference_model\n    prepend_feed_ops(main_program, feeded_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 0
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_0.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_0.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_0.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_1.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_1.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_1.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_2.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_2.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_2.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_3.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_3.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_3.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_4.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_4.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_4.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_5.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_5.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_5.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_6.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_6.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_6.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_7.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_7.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_7.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_8.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_8.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_8.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_9.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_9.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_9.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_10.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_10.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_10.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_11.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_11.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_11.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_12.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_12.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_12.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_13.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_13.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_13.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_14.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_14.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_14.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_15.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_15.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_15.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_16.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_16.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_16.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_17.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_17.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_17.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_18.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_18.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_18.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_19.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_19.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_19.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_20.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_20.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_20.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_21.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_21.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_21.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_22.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_22.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_22.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_23.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_23.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_23.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_24.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_24.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_24.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "embedding_25.tmp_0"
+    }
+    outputs {
+      parameter: "MaxIndex"
+      arguments: "sequence_pool_25.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "sequence_pool_25.tmp_0"
+    }
+    type: "sequence_pool"
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "pooltype"
+      type: STRING
+      s: "AVERAGE"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 2318, in sequence_pool\n    \"is_test\": is_test})\n"
+      strings: "  File \"save_program.py\", line 72, in embedding_layer\n    seq = fluid.layers.sequence_pool(input=emb, pool_type=\'average\')\n"
+      strings: "  File \"save_program.py\", line 86, in ctr_dnn_model\n    sparse_embed_and_seq = list(map(embedding_layer, words[1:-1]))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "sequence_pool_0.tmp_0"
+      arguments: "sequence_pool_1.tmp_0"
+      arguments: "sequence_pool_2.tmp_0"
+      arguments: "sequence_pool_3.tmp_0"
+      arguments: "sequence_pool_4.tmp_0"
+      arguments: "sequence_pool_5.tmp_0"
+      arguments: "sequence_pool_6.tmp_0"
+      arguments: "sequence_pool_7.tmp_0"
+      arguments: "sequence_pool_8.tmp_0"
+      arguments: "sequence_pool_9.tmp_0"
+      arguments: "sequence_pool_10.tmp_0"
+      arguments: "sequence_pool_11.tmp_0"
+      arguments: "sequence_pool_12.tmp_0"
+      arguments: "sequence_pool_13.tmp_0"
+      arguments: "sequence_pool_14.tmp_0"
+      arguments: "sequence_pool_15.tmp_0"
+      arguments: "sequence_pool_16.tmp_0"
+      arguments: "sequence_pool_17.tmp_0"
+      arguments: "sequence_pool_18.tmp_0"
+      arguments: "sequence_pool_19.tmp_0"
+      arguments: "sequence_pool_20.tmp_0"
+      arguments: "sequence_pool_21.tmp_0"
+      arguments: "sequence_pool_22.tmp_0"
+      arguments: "sequence_pool_23.tmp_0"
+      arguments: "sequence_pool_24.tmp_0"
+      arguments: "sequence_pool_25.tmp_0"
+      arguments: "dense_input"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "concat_0.tmp_0"
+    }
+    type: "concat"
+    attrs {
+      name: "axis"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/tensor.py\", line 214, in concat\n    attrs={\'axis\': axis})\n"
+      strings: "  File \"save_program.py\", line 90, in ctr_dnn_model\n    concated = fluid.layers.concat(sparse_embed_seq + words[0:1], axis=1)\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "use_mkldnn"
+      type: BOOLEAN
+      b: false
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "concat_0.tmp_0"
+    }
+    inputs {
+      parameter: "Y"
+      arguments: "fc_0.w_0"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_0.tmp_0"
+    }
+    type: "mul"
+    attrs {
+      name: "x_num_col_dims"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "y_num_col_dims"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 323, in fc\n    \"y_num_col_dims\": 1})\n"
+      strings: "  File \"save_program.py\", line 97, in ctr_dnn_model\n    scale=1 / math.sqrt(concated.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_0.tmp_0"
+    }
+    inputs {
+      parameter: "Y"
+      arguments: "fc_0.b_0"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_0.tmp_1"
+    }
+    type: "elementwise_add"
+    attrs {
+      name: "axis"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 135, in append_bias_op\n    attrs={\'axis\': dim_start})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 336, in fc\n    pre_activation = helper.append_bias_op(pre_bias, dim_start=num_flatten_dims)\n"
+      strings: "  File \"save_program.py\", line 97, in ctr_dnn_model\n    scale=1 / math.sqrt(concated.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "use_mkldnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "x_data_format"
+      type: STRING
+      s: ""
+    }
+    attrs {
+      name: "y_data_format"
+      type: STRING
+      s: ""
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_0.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_0.tmp_2"
+    }
+    type: "relu"
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 159, in append_activation\n    attrs=act)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 338, in fc\n    return helper.append_activation(pre_activation)\n"
+      strings: "  File \"save_program.py\", line 97, in ctr_dnn_model\n    scale=1 / math.sqrt(concated.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "use_mkldnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "use_cudnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_0.tmp_2"
+    }
+    inputs {
+      parameter: "Y"
+      arguments: "fc_1.w_0"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_1.tmp_0"
+    }
+    type: "mul"
+    attrs {
+      name: "x_num_col_dims"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "y_num_col_dims"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 323, in fc\n    \"y_num_col_dims\": 1})\n"
+      strings: "  File \"save_program.py\", line 100, in ctr_dnn_model\n    scale=1 / math.sqrt(fc1.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_1.tmp_0"
+    }
+    inputs {
+      parameter: "Y"
+      arguments: "fc_1.b_0"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_1.tmp_1"
+    }
+    type: "elementwise_add"
+    attrs {
+      name: "axis"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 135, in append_bias_op\n    attrs={\'axis\': dim_start})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 336, in fc\n    pre_activation = helper.append_bias_op(pre_bias, dim_start=num_flatten_dims)\n"
+      strings: "  File \"save_program.py\", line 100, in ctr_dnn_model\n    scale=1 / math.sqrt(fc1.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "use_mkldnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "x_data_format"
+      type: STRING
+      s: ""
+    }
+    attrs {
+      name: "y_data_format"
+      type: STRING
+      s: ""
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_1.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_1.tmp_2"
+    }
+    type: "relu"
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 159, in append_activation\n    attrs=act)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 338, in fc\n    return helper.append_activation(pre_activation)\n"
+      strings: "  File \"save_program.py\", line 100, in ctr_dnn_model\n    scale=1 / math.sqrt(fc1.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "use_mkldnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "use_cudnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_1.tmp_2"
+    }
+    inputs {
+      parameter: "Y"
+      arguments: "fc_2.w_0"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_2.tmp_0"
+    }
+    type: "mul"
+    attrs {
+      name: "x_num_col_dims"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "y_num_col_dims"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 323, in fc\n    \"y_num_col_dims\": 1})\n"
+      strings: "  File \"save_program.py\", line 103, in ctr_dnn_model\n    scale=1 / math.sqrt(fc2.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_2.tmp_0"
+    }
+    inputs {
+      parameter: "Y"
+      arguments: "fc_2.b_0"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_2.tmp_1"
+    }
+    type: "elementwise_add"
+    attrs {
+      name: "axis"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 135, in append_bias_op\n    attrs={\'axis\': dim_start})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 336, in fc\n    pre_activation = helper.append_bias_op(pre_bias, dim_start=num_flatten_dims)\n"
+      strings: "  File \"save_program.py\", line 103, in ctr_dnn_model\n    scale=1 / math.sqrt(fc2.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "use_mkldnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "x_data_format"
+      type: STRING
+      s: ""
+    }
+    attrs {
+      name: "y_data_format"
+      type: STRING
+      s: ""
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_2.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_2.tmp_2"
+    }
+    type: "relu"
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 159, in append_activation\n    attrs=act)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 338, in fc\n    return helper.append_activation(pre_activation)\n"
+      strings: "  File \"save_program.py\", line 103, in ctr_dnn_model\n    scale=1 / math.sqrt(fc2.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "use_mkldnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "use_cudnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_2.tmp_2"
+    }
+    inputs {
+      parameter: "Y"
+      arguments: "fc_3.w_0"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_3.tmp_0"
+    }
+    type: "mul"
+    attrs {
+      name: "x_num_col_dims"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "y_num_col_dims"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 323, in fc\n    \"y_num_col_dims\": 1})\n"
+      strings: "  File \"save_program.py\", line 106, in ctr_dnn_model\n    scale=1 / math.sqrt(fc3.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_3.tmp_0"
+    }
+    inputs {
+      parameter: "Y"
+      arguments: "fc_3.b_0"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_3.tmp_1"
+    }
+    type: "elementwise_add"
+    attrs {
+      name: "axis"
+      type: INT
+      i: 1
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 135, in append_bias_op\n    attrs={\'axis\': dim_start})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 336, in fc\n    pre_activation = helper.append_bias_op(pre_bias, dim_start=num_flatten_dims)\n"
+      strings: "  File \"save_program.py\", line 106, in ctr_dnn_model\n    scale=1 / math.sqrt(fc3.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "use_mkldnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "x_data_format"
+      type: STRING
+      s: ""
+    }
+    attrs {
+      name: "y_data_format"
+      type: STRING
+      s: ""
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_3.tmp_1"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fc_3.tmp_2"
+    }
+    type: "softmax"
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 159, in append_activation\n    attrs=act)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 338, in fc\n    return helper.append_activation(pre_activation)\n"
+      strings: "  File \"save_program.py\", line 106, in ctr_dnn_model\n    scale=1 / math.sqrt(fc3.shape[1]))))\n"
+      strings: "  File \"save_program.py\", line 156, in save_program\n    loss, auc_var, batch_auc_var, train_feed_vars, inference_feed_vars, fetch_vars = ctr_dnn_model(args.embedding_size, args.sparse_feature_dim, use_py_reader=False)\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "axis"
+      type: INT
+      i: -1
+    }
+    attrs {
+      name: "use_cudnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "data_format"
+      type: STRING
+      s: "AnyLayout"
+    }
+    attrs {
+      name: "use_mkldnn"
+      type: BOOLEAN
+      b: false
+    }
+    attrs {
+      name: "is_test"
+      type: INT
+      i: 1
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "fc_3.tmp_2"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "save_infer_model/scale_0"
+    }
+    type: "scale"
+    attrs {
+      name: "scale"
+      type: FLOAT
+      f: 1.0
+    }
+    attrs {
+      name: "bias"
+      type: FLOAT
+      f: 0.0
+    }
+    attrs {
+      name: "bias_after_scale"
+      type: BOOLEAN
+      b: true
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layer_helper.py\", line 43, in append_op\n    return self.main_program.current_block().append_op(*args, **kwargs)\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/layers/nn.py\", line 9267, in scale\n    \'bias_after_scale\': bias_after_scale\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 957, in save_inference_model\n    var, 1., name=\"save_infer_model/scale_{}\".format(i))\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+  }
+  ops {
+    inputs {
+      parameter: "X"
+      arguments: "save_infer_model/scale_0"
+    }
+    outputs {
+      parameter: "Out"
+      arguments: "fetch"
+    }
+    type: "fetch"
+    attrs {
+      name: "op_role_var"
+      type: STRINGS
+    }
+    attrs {
+      name: "op_callstack"
+      type: STRINGS
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/framework.py\", line 1654, in append_op\n    attrs=kwargs.get(\"attrs\", None))\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 862, in append_fetch_ops\n    attrs={\'col\': i})\n"
+      strings: "  File \"/home/wangguibao/paddle_env/paddle_release_home/python/lib/python2.7/site-packages/paddle/fluid/io.py\", line 1001, in save_inference_model\n    append_fetch_ops(main_program, fetch_var_names)\n"
+      strings: "  File \"save_program.py\", line 168, in save_program\n    fluid.io.save_inference_model(model_dir, feed_var_names, fetch_vars, exe, fluid.default_main_program())\n"
+      strings: "  File \"save_program.py\", line 195, in <module>\n    save_program()\n"
+    }
+    attrs {
+      name: "op_namescope"
+      type: STRING
+      s: "/"
+    }
+    attrs {
+      name: "op_role"
+      type: INT
+      i: 0
+    }
+    attrs {
+      name: "col"
+      type: INT
+      i: 0
+    }
+  }
+}