peft/pr_3206/en/package_reference/helpers.md

# Helper methods

A collection of helper functions for PEFT.

## Checking if a model is a PEFT model[[peft.helpers.check_if_peft_model]]

#### peft.helpers.check_if_peft_model[[peft.helpers.check_if_peft_model]]

[Source](https://github.com/huggingface/peft/blob/vr_3206/src/peft/helpers.py#L135)

Check if the model is a PEFT model.

**Parameters:**

model_name_or_path (`str`) : Model id to check, can be local or on the Hugging Face Hub.

**Returns:**

``bool``

True if the model is a PEFT model, False otherwise.

## Temporarily Rescaling Adapter Scale in LoraLayer Modules[[peft.helpers.rescale_adapter_scale]]

#### peft.helpers.rescale_adapter_scale[[peft.helpers.rescale_adapter_scale]]

[Source](https://github.com/huggingface/peft/blob/vr_3206/src/peft/helpers.py#L156)

Context manager to temporarily rescale the scaling of the LoRA adapter in a model.

The original scaling values are restored when the context manager exits. This context manager works with the
transformers and diffusers models that have directly loaded LoRA adapters.

For LoRA, applying this context manager with multiplier in [0, 1] is strictly equivalent to applying
[wise-ft](https://huggingface.co/papers/2109.01903) (see [#1940](https://github.com/huggingface/peft/issues/1940)
for details). It can improve the performances of the model if there is a distribution shiftbetween the training
data used for fine-tuning, and the test data used during inference.

Warning: It has been reported that when using Apple's MPS backend for PyTorch, it is necessary to add a short sleep
time after exiting the context before the scales are fully restored.

Example:

```python
>>> model = ModelWithLoraLayer()
>>> multiplier = 0.5
>>> with rescale_adapter_scale(model, multiplier):
...     outputs = model(**inputs)  # Perform operations with the scaled model
>>> outputs = model(**inputs)  # The original scaling values are restored here
```

**Parameters:**

model : The model containing `LoraLayer` modules whose scaling is to be adjusted.

multiplier (float or int) : The multiplier that rescales the `scaling` attribute. Must be of type float or int.

## Context manager to disable input dtype casting in the `forward` method of LoRA layers[[peft.helpers.disable_input_dtype_casting]]

#### peft.helpers.disable_input_dtype_casting[[peft.helpers.disable_input_dtype_casting]]

[Source](https://github.com/huggingface/peft/blob/vr_3206/src/peft/helpers.py#L217)

Context manager disables input dtype casting to the dtype of the weight.

**Parameters:**

model (nn.Module) : The model containing PEFT modules whose input dtype casting is to be adjusted.

active (bool) : Whether the context manager is active (default) or inactive.

## Context manager to enable DoRA caching (faster at inference time but requires more memory)[[peft.helpers.DoraCaching]]

#### peft.helpers.DoraCaching[[peft.helpers.DoraCaching]]

[Source](https://github.com/huggingface/peft/blob/vr_3206/src/peft/helpers.py#L254)

Context manager to enable DoRA caching, which improves speed of DoRA inference at the expense of memory.

With active caching, the materialized LoRA weight (B @ A) and the weight norm (base weight + LoRA weight) are
cached.

Even within the caching context, if the model is in training mode, caching is disabled. When the model switches to
training mode, the cache will be cleared.

Example:

```py
>>> from peft.helpers import enable_dora_scaling

>>> model.eval()  # put in eval model for caching to work

>>> with DoraCaching():  # use as a context manager
...     output = model(inputs)

>>> dora_caching = DoraCaching()
>>> dora_caching(enabled=True)  # permanently enable caching
>>> output = model(inputs)
>>> dora_caching(enabled=False)  # permanently disable caching
>>> output = model(inputs)
```