memcpy(&product[0], TF_TensorData(t), TF_TensorByteSize(t));
  TF_DeleteTensor(t);
  EXPECT_EQ(7, product[0]);
  EXPECT_EQ(10, product[1]);
  EXPECT_EQ(15, product[2]);
  EXPECT_EQ(22, product[3]);

  TFE_DeleteTensorHandle(h0_task0);
  TFE_DeleteTensorHandle(h1_task0);
  if (packed_handle) {
    TFE_DeleteTensorHandle(packed_handle);
  }
  TFE_DeleteTensorHandle(h1_task2);

    style V1Field fill:#fff6f0
    style V2Field fill:#f9fff3
```

어떤 경우에는 FastAPI 앱의 같은 **경로 처리**에서 Pydantic v1과 v2 모델을 함께 사용하는 것도 가능합니다:

{* ../../docs_src/pydantic_v1_in_v2/tutorial003_an_py310.py hl[2:3,6,12,21:22] *}

위 예제에서 입력 모델은 Pydantic v1 모델이고, 출력 모델(`response_model=ItemV2`로 정의됨)은 Pydantic v2 모델입니다.

### Pydantic v1 파라미터 { #pydantic-v1-parameters }

```

In some cases, it's even possible to have both Pydantic v1 and v2 models in the same **path operation** in your FastAPI app:

{* ../../docs_src/pydantic_v1_in_v2/tutorial003_an_py310.py hl[2:3,6,12,21:22] *}

In this example above, the input model is a Pydantic v1 model, and the output model (defined in `response_model=ItemV2`) is a Pydantic v2 model.

### Pydantic v1 parameters { #pydantic-v1-parameters }

igarashi <******@****.***> 1565681654 +0900

    }

    private
    fun getMergedContributorPullRequests(pageNumber: Int): List<GitHubPullRequest> {
        val uri = "https://api.github.com/search/issues?q=is:pr+is:merged+repo:gradle/gradle+label:%22from:contributor%22&sort=updated&order=desc&per_page=$PAGE_SIZE&page=$pageNumber"
        return invokeGitHubApi(uri, GitHubPullRequestSearchResult::class.java).items
    }

{* ../../docs_src/response_model/tutorial003_py310.py hl[24] *}

...we declared the `response_model` to be our model `UserOut`, that doesn't include the password:

{* ../../docs_src/response_model/tutorial003_py310.py hl[22] *}

So, **FastAPI** will take care of filtering out all the data that is not declared in the output model (using Pydantic).

### `response_model` or Return Type { #response-model-or-return-type }

    style V1Field fill:#fff6f0
    style V2Field fill:#f9fff3
```

在某些情况下，甚至可以在 FastAPI 应用的同一个路径操作中同时使用 Pydantic v1 和 v2 模型：

{* ../../docs_src/pydantic_v1_in_v2/tutorial003_an_py310.py hl[2:3,6,12,21:22] *}

在上面的示例中，输入模型是 Pydantic v1 模型，输出模型（在 `response_model=ItemV2` 中定义）是 Pydantic v2 模型。

### Pydantic v1 参数 { #pydantic-v1-parameters }

当你打开自动 API 文档时，你的扩展会显示在该 *路径操作* 的底部。

<img src="/img/tutorial/path-operation-advanced-configuration/image01.png">

如果你查看最终生成的 OpenAPI（在你的 API 的 `/openapi.json`），你也会看到你的扩展作为该 *路径操作* 的一部分：

```JSON hl_lines="22"
{
    "openapi": "3.1.0",
    "info": {
        "title": "FastAPI",
        "version": "0.1.0"
    },
    "paths": {
        "/items/": {
            "get": {

That way, we can declare just the differences between the models (with plaintext `password`, with `hashed_password` and without password):

{* ../../docs_src/extra_models/tutorial002_py310.py hl[7,13:14,17:18,21:22] *}

## `Union` or `anyOf` { #union-or-anyof }

You can declare a response to be the `Union` of two or more types, that means, that the response would be any of them.

It will be defined in OpenAPI with `anyOf`.

            (byte) 26, (byte) 18, (byte) 10, (byte) 2, (byte) 59, (byte) 51, (byte) 43, (byte) 35, (byte) 62, (byte) 54, (byte) 46,
            (byte) 38, (byte) 30, (byte) 22, (byte) 14, (byte) 6, (byte) 61, (byte) 53, (byte) 45, (byte) 37, (byte) 29, (byte) 21,
            (byte) 13, (byte) 5, (byte) 60, (byte) 52, (byte) 44, (byte) 36, (byte) 28, (byte) 20, (byte) 12, (byte) 4, (byte) 27,