# Utilidades generales

Esta página enumera todas las funciones de utilidad generales de Transformers que se encuentran en el archivo `utils.py`.

La mayoría de ellos solo son útiles si estás estudiando el código general de la biblioteca.

## Enums y namedtuples[[transformers.utils.ExplicitEnum]]

#### transformers.utils.ExplicitEnum[[transformers.utils.ExplicitEnum]]

```python
transformers.utils.ExplicitEnum(value, names = None, module = None, qualname = None, type = None, start = 1, boundary = None)
```

[Fuente](https://github.com/huggingface/transformers/blob/main/src/transformers/utils/generic.py#L571)

Enum con un mensaje de error más explícito para valores faltantes.

#### transformers.utils.PaddingStrategy[[transformers.utils.PaddingStrategy]]

```python
transformers.utils.PaddingStrategy(value, names = None, module = None, qualname = None, type = None, start = 1, boundary = None)
```

[Fuente](https://github.com/huggingface/transformers/blob/main/src/transformers/utils/generic.py#L583)

Valores posibles para el argumento `padding` en [PreTrainedTokenizerBase.__call__()](/docs/transformers/main/es/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__). Útil para la autocompletado de pestañas en un IDE.

#### transformers.TensorType[[transformers.TensorType]]

```python
transformers.TensorType(value, names = None, module = None, qualname = None, type = None, start = 1, boundary = None)
```

[Fuente](https://github.com/huggingface/transformers/blob/main/src/transformers/utils/generic.py#L594)

Valores posibles para el argumento `return_tensors` en [PreTrainedTokenizerBase.__call__()](/docs/transformers/main/es/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__). Útil para la autocompletado de pestañas en un IDE.

## Decoradores especiales[[transformers.add_start_docstrings]]

#### transformers.add_start_docstrings[[transformers.add_start_docstrings]]

```python
transformers.add_start_docstrings(*docstr)
```

[Fuente](https://github.com/huggingface/transformers/blob/main/src/transformers/utils/doc.py#L38)

#### transformers.utils.add_start_docstrings_to_model_forward[[transformers.utils.add_start_docstrings_to_model_forward]]

```python
transformers.utils.add_start_docstrings_to_model_forward(*docstr)
```

[Fuente](https://github.com/huggingface/transformers/blob/main/src/transformers/utils/doc.py#L46)

#### transformers.add_end_docstrings[[transformers.add_end_docstrings]]

```python
transformers.add_end_docstrings(*docstr)
```

[Fuente](https://github.com/huggingface/transformers/blob/main/src/transformers/utils/doc.py#L82)

#### transformers.utils.add_code_sample_docstrings[[transformers.utils.add_code_sample_docstrings]]

```python
transformers.utils.add_code_sample_docstrings(*docstr, processor_class = None, checkpoint = None, output_type = None, config_class = None, mask = '[MASK]', qa_target_start_index = 14, qa_target_end_index = 15, model_cls = None, modality = None, expected_output = None, expected_loss = None, real_checkpoint = None, revision = None)
```

[Fuente](https://github.com/huggingface/transformers/blob/main/src/transformers/utils/doc.py#L969)

#### transformers.utils.replace_return_docstrings[[transformers.utils.replace_return_docstrings]]

```python
transformers.utils.replace_return_docstrings(output_type = None, config_class = None)
```

[Fuente](https://github.com/huggingface/transformers/blob/main/src/transformers/utils/doc.py#L1063)

## Otras utilidades[[transformers.utils._LazyModule]]

#### transformers.utils._LazyModule[[transformers.utils._LazyModule]]

```python
transformers.utils._LazyModule(name: str, module_file: str, import_structure: dict, module_spec: _frozen_importlib.ModuleSpec | None = None, extra_objects: dict[str, object] | None = None, explicit_import_shortcut: dict[str, list[str]] | None = None)
```

[Fuente](https://github.com/huggingface/transformers/blob/main/src/transformers/utils/import_utils.py#L2499)

Clase de módulo que expone todos los objetos, pero solo realiza las importaciones asociadas cuando se solicitan los objetos.