А `dict`, с именем `weights`, который вы получите в качестве ответа Pydantic, действительно будет иметь ключи типа `int` и значения типа `float`.

## Резюме

    This means that, even though your API clients can only send strings as keys, as long as those strings contain pure integers, Pydantic will convert them and validate them.

    And the `dict` you receive as `weights` will actually have `int` keys and `float` values.

## Recap

With **FastAPI** you have the maximum flexibility provided by Pydantic models, while keeping your code simple, short and elegant.

    Das bedeutet, dass Ihre API-Clients nur Strings senden können, aber solange diese Strings nur Zahlen enthalten, wird Pydantic sie konvertieren und validieren.

    Und das `dict` welches Sie als `weights` erhalten, wird `int`-Schlüssel und `float`-Werte haben.

## Zusammenfassung

Mit **FastAPI** haben Sie die maximale Flexibilität von Pydantic-Modellen, während Ihr Code einfach, kurz und elegant bleibt.

                    type.getNumElements());
  }

  return sparsity;
}

typedef struct InspectResult {
  // Whether the weight tensor is sparse enough to be compressed.
  bool can_compress;
  // If the weight tensor cannot be encoded in a block configuration that the op
  // supports, a Densify() op will be inserted afterwards to fall back to dense
  // execution.
  bool needs_densify;

		// encountered, let the individual Score weight take precedence.
		if _, ok := f.scorePluginWeight[e.Name]; ok {
			continue
		}
		// a weight of zero is not permitted, plugins can be disabled explicitly
		// when configured.
		f.scorePluginWeight[e.Name] = int(e.Weight)
		if f.scorePluginWeight[e.Name] == 0 {
			f.scorePluginWeight[e.Name] = 1
		}

// resource ops are considered duplicated if they have the same `shared_name`.
std::unique_ptr<OperationPass<func::FuncOp>>
CreateMergeDuplicateResourceOpsPass();

// Apply quantization to weights based on the provided schemes.
std::unique_ptr<OperationPass<ModuleOp>> CreateQuantizeWeightsPass(
    const tensorflow::quantization::QuantizationOptions& quant_options);

// Propagate quantized type through allowed ops.

    ```

!!! tip
    请记住 JSON 仅支持将 `str` 作为键。

    但是 Pydantic 具有自动转换数据的功能。

    这意味着，即使你的 API 客户端只能将字符串作为键发送，只要这些字符串内容仅包含整数，Pydantic 就会对其进行转换并校验。

    然后你接收的名为 `weights` 的 `dict` 实际上将具有 `int` 类型的键和 `float` 类型的值。

## 总结

使用 **FastAPI** 你可以拥有 Pydantic 模型提供的极高灵活性，同时保持代码的简单、简短和优雅。

而且还具有下列好处：

* 编辑器支持（处处皆可自动补全！）
* 数据转换（也被称为解析/序列化）
* 数据校验
* 模式文档

//       value {dimension_specs {dimension: 3}}}
//   }}
// }
// ```
//
// This preset:
//   * Applies per-channel quantization for weights (input index 1) of
//     convolution quantizable unit family. The quantization dimension is 3, the
//     channel dimension, which assumes the weight tensor is in NHWC format.
//   * Applies static-range PTQ for all other ops.
//
// Next ID: 4
message StaticRangePtqPreset {

constexpr mlir::StringRef kUnidirectionalSequenceRnnOp =
    "name: 'UnidirectionalSequenceRnn' input_arg: {name: 'Input' type: "
    "DT_FLOAT} input_arg: { name: 'Weights' type: DT_FLOAT } "
    "input_arg: { name: 'RecurrentWeights' type: DT_FLOAT } input_arg: { "
    "name: 'Bias' type: DT_FLOAT} "
    "input_arg: { name: 'HiddenState' type: DT_FLOAT} "
    "output_arg: { name: "

	}
	return
}

func validateRouteDestinations(weights []*networking.RouteDestination, gatewaySemantics bool) (errs error) {
	var totalWeight int32
	for _, weight := range weights {
		if weight == nil {
			errs = multierror.Append(errs, errors.New("weight may not be nil"))
			continue
		}
		if weight.Destination == nil {