pyiqa.models.inference_model
============================

.. py:module:: pyiqa.models.inference_model




Module Contents
---------------

.. py:class:: InferenceModel(metric_name, as_loss=False, loss_weight=None, loss_reduction='mean', device=None, seed=123, check_input_range=True, **kwargs)

   Bases: :py:obj:`torch.nn.Module`


   Common interface for quality inference of images with default setting of each metric.


   .. py:method:: load_weights(weights_path, weight_keys='params')


   .. py:method:: is_valid_input(x)


   .. py:method:: forward(target, ref=None, **kwargs)