tensor<i32> {tf._user_specified_name = "args_3"}, %arg4: tensor<?x!tf_type.string> {tf._user_specified_name = "args_4"}, %arg5: tensor<?x!tf_type.string> {tf._user_specified_name = "args_5"}, %arg6: tensor<?x!tf_type.string> {tf._user_specified_name = "args_6"}, %arg7: tensor<?x?x512xf32> {tf._user_specified_name = "args_7"}, %arg8: tensor<?x?xf32> {tf._user_specified_name = "args_8"}, %arg9: tensor<?x300xi32> {tf._user_specified_name = "args_9"}, %arg10: tensor<?x300xi32> {tf._user_specified_name = "args_10"},...

  // CHECK-SAME: (%[[ARG_0:[a-z0-9]*]]: tensor<*xi32>, %[[ARG_1:[a-z0-9]*]]: tensor<8xi32>, %[[ARG_2:[a-z0-9]*]]: tensor<*xi32>, %[[ARG_3:[a-z0-9]*]]: tensor<8xi32>)
  func.func @static_and_dynamic_shapes(%arg0: tensor<*xi32>, %arg1: tensor<8xi32>, %arg2: tensor<*xi32>, %arg3: tensor<8xi32>) -> tensor<8xi32> {
    // CHECK-NOT:  "tf.Shape"(%[[ARG_1]])
    // CHECK-NOT:  "tf.Shape"(%[[ARG_3]])

      func.return
    }
    ```

    with the following reduction function:

    ```mlir
    func.func private @__reduce_func_1(%arg0: tensor<i64> {tf._user_specified_name = "args_0"},
      %arg1: tensor<32xf32> {tf._user_specified_name = "args_1"}) -> (tensor<i64>)
      attributes {tf._tf_data_function = true, tf.signature.is_stateful} {
        %0 = "tf.JustPretend"(%arg1) : (tensor<32xf32>) -> (tensor<i64>)

}