"get": {
                    "responses": {
                        "200": {
                            "description": "Successful Response",
                            "content": {"application/json": {"schema": {}}},
                        }
                    },
                    "summary": "Read Items",
                    "operationId": "read_items_items__get",
                }
            }

                        "responses": {
                            "200": {
                                "description": "Successful Response",
                                "content": {
                                    "application/json": {
                                        "schema": {
                                            "$ref": "#/components/schemas/ModelWithRef"
                                        }

purpose GPUs, and custom-designed ASICs known as Tensor Processing Units (TPUs). This architecture gives flexibility to the application developer, whereas in previous “parameter server” designs the management of shared state is built into the system, TensorFlow enables developers to experiment with novel optimizations and training algorithms. TensorFlow supports a variety of applications, with a focus on training and inference on deep neural networks. Several Google services use TensorFlow in production,...

Das ist keine Limitation von **FastAPI**, sondern Teil des HTTP-Protokolls.

///

## Zusammenfassung { #recap }

但如果你直接返回 `Response`，返回数据不会自动转换，也不会自动生成文档（例如，在 HTTP 头 `Content-Type` 中包含特定的「媒体类型」作为生成的 OpenAPI 的一部分）。

你还可以在 *路径操作装饰器* 中声明你想用的 `Response`。

你从 *路径操作函数* 中返回的内容将被放在该 `Response` 中。

并且如果该 `Response` 有一个 JSON 媒体类型（`application/json`），比如使用 `JSONResponse` 或者 `UJSONResponse` 的时候，返回的数据将使用你在路径操作装饰器中声明的任何 Pydantic 的 `response_model` 自动转换（和过滤）。

/// note | 说明

如果你使用不带有任何媒体类型的响应类，FastAPI 认为你的响应没有任何内容，所以不会在生成的OpenAPI文档中记录响应格式。

///

                "get": {
                    "responses": {
                        "200": {
                            "description": "Successful Response",
                            "content": {"application/json": {"schema": {}}},
                        }
                    },
                    "summary": "Read Current User",
                    "operationId": "read_current_user_users_me_get",

# Low latency mode for TCP
net.ipv4.tcp_low_latency=1

# The following variable is used to tell the kernel how 
# much of the socket buffer space should be used for TCP 
# window size, and how much to save for an application buffer.
net.ipv4.tcp_adv_win_scale=1

# disable RFC2861 behavior
net.ipv4.tcp_slow_start_after_idle = 0

# Fix faulty network setups
net.ipv4.tcp_mtu_probing=1
net.ipv4.tcp_base_mss=1280

# Disable ipv6

worker_processes  auto;

error_log  /var/log/nginx/error.log warn;
pid        /var/run/nginx.pid;

events {
    worker_connections  4096;
}

http {
    include       /etc/nginx/mime.types;
    default_type  application/octet-stream;

    log_format  main  '$remote_addr - $remote_user [$time_local] "$request" '
                      '$status $body_bytes_sent "$http_referer" '
                      '"$http_user_agent" "$http_x_forwarded_for"';

    pluginEntriesFrom(file)
        .asSequence()
        .map { (id, implementationClass) ->
            val simpleId = id.substringAfter("org.gradle.")
            // One plugin extension for the simple id, e.g., "application"
            PluginExtension(simpleId, id, implementationClass)

    n = 1000
    headers = {}
    body = [1] * n
    data = json.dumps(body).encode()
    if compress:
        data = gzip.compress(data)
        headers["Content-Encoding"] = "gzip"
    headers["Content-Type"] = "application/json"
    response = client.post("/sum", content=data, headers=headers)
    assert response.json() == {"sum": n}


def test_request_class(client: TestClient):
    response = client.get("/check-class")