"input": "foo",
                    "ctx": {"expected": "'alexnet', 'resnet' or 'lenet'"},
                }
            ]
        }
    ) | IsDict(
        # TODO: remove when deprecating Pydantic v1
        {
            "detail": [
                {
                    "ctx": {"enum_values": ["alexnet", "resnet", "lenet"]},

from enum import Enum

from fastapi import FastAPI


class ModelName(str, Enum):
    alexnet = "alexnet"
    resnet = "resnet"
    lenet = "lenet"


app = FastAPI()


@app.get("/models/{model_name}")
async def get_model(model_name: ModelName):
    if model_name is ModelName.alexnet:
        return {"model_name": model_name, "message": "Deep Learning FTW!"}

    if model_name.value == "lenet":

<a href="https://docs.python.org/3/library/enum.html" class="external-link" target="_blank">Enumerations (or enums) are available in Python</a> since version 3.4.

///

/// tip

If you are wondering, "AlexNet", "ResNet", and "LeNet" are just names of Machine Learning <abbr title="Technically, Deep Learning model architectures">models</abbr>.

///

### Declare a *path parameter*

///

/// tip | Consejo

Si lo estás dudando, "AlexNet", "ResNet", y "LeNet" son solo nombres de <abbr title="Técnicamente, arquitecturas de modelos de Deep Learning">modelos</abbr> de Machine Learning.

///

### Declara un *parámetro de path*

```

/// info | "说明"

Python 3.4 及之后版本支持<a href="https://docs.python.org/3/library/enum.html" class="external-link" target="_blank">枚举（即 enums）</a>。

///

/// tip | "提示"

**AlexNet**、**ResNet**、**LeNet** 是机器学习<abbr title="技术上来说是深度学习模型架构">模型</abbr>。

///

### 声明*路径参数*

使用 Enum 类（`ModelName`）创建使用类型注解的*路径参数*：

```Python hl_lines="16"
{!../../docs_src/path_params/tutorial005.py!}
```

    assertEquals("9.999 \u03bcs", stopwatch.toString());
    stopwatch.reset();
    stopwatch.start();
    ticker.advance(1234567);
    assertEquals("1.235 ms", stopwatch.toString());
    stopwatch.reset();
    stopwatch.start();
    ticker.advance(5000000000L);
    assertEquals("5.000 s", stopwatch.toString());
    stopwatch.reset();
    stopwatch.start();
    ticker.advance((long) (1.5 * 60 * 1000000000L));

    assertTrue(reader.markSupported());

    assertEquals(string, readFully(reader));
    assertFullyRead(reader);

    // reset and read again
    reader.reset();
    assertEquals(string, readFully(reader));
    assertFullyRead(reader);

    // reset, skip, mark, then read the rest
    reader.reset();
    assertEquals(5, reader.skip(5));
    reader.mark(Integer.MAX_VALUE);
    assertEquals(string.substring(5), readFully(reader));

    A.reset(A_raw);
  }
  float B_vals[] = {.5f, -1.0f, 1.0f, 1.0f};
  int64_t B_dims[] = {2, 2};
  AbstractTensorHandlePtr B;
  {
    AbstractTensorHandle* B_raw;
    absl::Status s = TestTensorHandleWithDims<float, TF_FLOAT>(
        ctx_.get(), B_vals, B_dims, 2, &B_raw);
    ASSERT_EQ(errors::OK, s.code()) << s.message();
    B.reset(B_raw);
  }

    assertEquals("9.999 \u03bcs", stopwatch.toString());
    stopwatch.reset();
    stopwatch.start();
    ticker.advance(1234567);
    assertEquals("1.235 ms", stopwatch.toString());
    stopwatch.reset();
    stopwatch.start();
    ticker.advance(5000000000L);
    assertEquals("5.000 s", stopwatch.toString());
    stopwatch.reset();
    stopwatch.start();
    ticker.advance((long) (1.5 * 60 * 1000000000L));

	)

	if config.Colorful {
		infoStr = Green + "%s\n" + Reset + Green + "[info] " + Reset
		warnStr = BlueBold + "%s\n" + Reset + Magenta + "[warn] " + Reset
		errStr = Magenta + "%s\n" + Reset + Red + "[error] " + Reset
		traceStr = Green + "%s\n" + Reset + Yellow + "[%.3fms] " + BlueBold + "[rows:%v]" + Reset + " %s"
		traceWarnStr = Green + "%s " + Yellow + "%s\n" + Reset + RedBold + "[%.3fms] " + Yellow + "[rows:%v]" + Magenta + " %s" + Reset