}

// Get the max throughput and iops numbers.
func objectSpeedTest(ctx context.Context, opts speedTestOpts) chan madmin.SpeedTestResult {
	ch := make(chan madmin.SpeedTestResult, 1)
	go func() {
		defer xioutil.SafeClose(ch)

		concurrency := opts.concurrencyStart

		if opts.autotune {
			// if we have less drives than concurrency then choose
			// only the concurrency to be number of drives to start

    ```

## Number validations: floats, greater than and less than

Number validations also work for `float` values.

Here's where it becomes important to be able to declare <abbr title="greater than"><code>gt</code></abbr> and not just <abbr title="greater than or equal"><code>ge</code></abbr>. As with it you can require, for example, that a value must be greater than `0`, even if it is less than `1`.

	UserTimeout        int              // this value is expected to be in milliseconds
	ClientReadTimeout  time.Duration    // When the net.Conn is idle for more than ReadTimeout duration, we close the connection on the client proactively.
	ClientWriteTimeout time.Duration    // When the net.Conn is idle for more than WriteTimeout duration, we close the connection on the client proactively.

//           value: image/* # match objects with 'content-type', all values starting with 'image/'
//       size:
//         lessThan: "10MiB" # match objects with size less than this value (e.g. 10MiB)
//         greaterThan: 1MiB # match objects with size greater than this value (e.g. 1MiB)
//       purge:
//           # retainVersions: 0 # (default) delete all versions of the object. This option is the fastest.

type BytePoolCap struct {
	c    chan []byte
	w    int
	wcap int
}

// NewBytePoolCap creates a new BytePool bounded to the given maxSize, with new
// byte arrays sized based on width.
func NewBytePoolCap(maxSize uint64, width int, capwidth int) (bp *BytePoolCap) {
	if capwidth > 0 && capwidth < 64 {
		panic("buffer capped with smaller than 64 bytes is not supported")
	}

E você também pode declarar validações numéricas:

* `gt`: maior que (`g`reater `t`han)
* `ge`: maior que ou igual (`g`reater than or `e`qual)
* `lt`: menor que (`l`ess `t`han)
* `le`: menor que ou igual (`l`ess than or `e`qual)

!!! info "Informação"
    `Query`, `Path` e outras classes que você verá a frente são subclasses de uma classe comum `Param`.

## 숫자 검증: 크거나 같음 및 작거나 같음

동일하게 적용됩니다:

* `gt`: 크거나(`g`reater `t`han)
* `le`: 작거나 같은(`l`ess than or `e`qual)

```Python hl_lines="9"
{!../../../docs_src/path_params_numeric_validations/tutorial005.py!}
```

## 숫자 검증: 부동소수, 크거나 및 작거나

숫자 검증은 `float` 값에도 동작합니다.

    *   Unary GPU kernels: Abs, Atanh, Acos, Acosh, Asin, Asinh, Atan, Cos, Cosh, Sin, Sinh, Tan, Tanh.
    *   Binary GPU kernels: AddV2, Sub, Div, DivNoNan, Mul, MulNoNan, FloorDiv, Equal, NotEqual, Greater, GreaterEqual, LessEqual, Less.

* `tf.lite`
    * Add experimental supports conversion of models that may be larger than 2GB before buffer deduplication

また、数値のバリデーションを宣言することもできます:

* `gt`: より大きい（`g`reater `t`han）
* `ge`: 以上（`g`reater than or `e`qual）
* `lt`: より小さい（`l`ess `t`han）
* `le`: 以下（`l`ess than or `e`qual）

!!! info "情報"
    `Query`、`Path`などは後に共通の`Param`クラスのサブクラスを見ることになります。（使う必要はありません）

    そして、それらすべては、これまで見てきた追加のバリデーションとメタデータと同じパラメータを共有しています。

А также вы можете добавить валидацию числовых данных:

* `gt`: больше (`g`reater `t`han)
* `ge`: больше или равно (`g`reater than or `e`qual)
* `lt`: меньше (`l`ess `t`han)
* `le`: меньше или равно (`l`ess than or `e`qual)

!!! info "Информация"
    `Query`, `Path` и другие классы, которые мы разберём позже, являются наследниками общего класса `Param`.