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

### Bug Fixes and Other Changes

* `tf.py_function` and `tf.numpy_function` can now be used as function decorators for clearer code:
   ```
   @tf.py_function(Tout=tf.float32)
   def my_fun(x):
     print("This always executes eagerly.")
     return x+1
   ```

* `tf.lite`

 * override one or more methods to modify the behavior of the backing iterator as desired per the <a
 * href="http://en.wikipedia.org/wiki/Decorator_pattern">decorator pattern</a>.
 *
 * <p><b>{@code default} method warning:</b> This class forwards calls to <i>only some</i> {@code
 * default} methods. Specifically, it forwards calls only for methods that existed <a

 * override one or more methods to modify the behavior of the backing map as desired per the <a
 * href="http://en.wikipedia.org/wiki/Decorator_pattern">decorator pattern</a>.
 *
 * <p><b>{@code default} method warning:</b> This class forwards calls to <i>only some</i> {@code
 * default} methods. Specifically, it forwards calls only for methods that existed <a

# Pfadoperation-Konfiguration

Es gibt mehrere Konfigurations-Parameter, die Sie Ihrem *Pfadoperation-Dekorator* übergeben können.

/// warning | Achtung

Beachten Sie, dass diese Parameter direkt dem *Pfadoperation-Dekorator* übergeben werden, nicht der *Pfadoperation-Funktion*.

///

## Response-Statuscode

Sie können den (HTTP-)`status_code` definieren, den die Response Ihrer *Pfadoperation* verwenden soll.

 * Subclasses should override one or more methods to modify the behavior of the backing collection
 * as desired per the <a href="http://en.wikipedia.org/wiki/Decorator_pattern">decorator
 * pattern</a>.
 *
 * <p><b>{@code default} method warning:</b> This class does <i>not</i> forward calls to {@code
 * default} methods. Instead, it inherits their default implementations. When those implementations

 * override one or more methods to modify the behavior of the backing multimap as desired per the <a
 * href="http://en.wikipedia.org/wiki/Decorator_pattern">decorator pattern</a>.
 *
 * <p><b>{@code default} method warning:</b> This class does <i>not</i> forward calls to {@code
 * default} methods. Instead, it inherits their default implementations. When those implementations

 * one or more methods to modify the behavior of the backing iterator as desired per the <a
 * href="http://en.wikipedia.org/wiki/Decorator_pattern">decorator pattern</a>.
 *
 * <p><b>{@code default} method warning:</b> This class forwards calls to <i>only some</i> {@code
 * default} methods. Specifically, it forwards calls only for methods that existed <a

 * should override one or more methods to modify the behavior of the backing executor service as
 * desired per the <a href="http://en.wikipedia.org/wiki/Decorator_pattern">decorator pattern</a>.
 *
 * <p><b>{@code default} method warning:</b> This class does <i>not</i> forward calls to {@code
 * default} methods. Instead, it inherits their default implementations. When those implementations

```Python
def get_settings():
    return Settings()
```

Iriamos criar um novo objeto a cada requisição, e estaríamos lendo o arquivo `.env` a cada requisição. ⚠️

Mas como estamos utilizando o decorador `@lru_cache` acima, o objeto `Settings` é criado apenas uma vez, na primeira vez que a função é chamada. ✔️

{* ../../docs_src/settings/app03_an_py39/main.py hl[1,11] *}

 * Subclasses should override one or more methods to modify the behavior of the backing collection
 * as desired per the <a href="http://en.wikipedia.org/wiki/Decorator_pattern">decorator
 * pattern</a>.
 *
 * <p><b>{@code default} method warning:</b> This class does <i>not</i> forward calls to {@code
 * default} methods. Instead, it inherits their default implementations. When those implementations