vision_unlearning.utils.gradient_weighting
==========================================

.. py:module:: vision_unlearning.utils.gradient_weighting


Attributes
----------

.. autoapisummary::

   vision_unlearning.utils.gradient_weighting.logger


Classes
-------

.. autoapisummary::

   vision_unlearning.utils.gradient_weighting.GradientWeightingMethod
   vision_unlearning.utils.gradient_weighting.GradientWeightingMethodNone
   vision_unlearning.utils.gradient_weighting.GradientWeightingMethodSimple
   vision_unlearning.utils.gradient_weighting.GradientWeightingMethodMunba


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

.. py:data:: logger

.. py:class:: GradientWeightingMethod(/, **data: Any)

   Bases: :py:obj:`pydantic.BaseModel`, :py:obj:`abc.ABC`


   Method used to conciliate/harmonize/combine/weight the gradients of the different tasks

   Inspired by @article{navon2022multi,
       title={Multi-Task Learning as a Bargaining Game},
       author={Navon, Aviv and Shamsian, Aviv and Achituve, Idan and Maron, Haggai and Kawaguchi, Kenji and Chechik, Gal and Fetaya, Ethan},
       journal={arXiv preprint arXiv:2202.01017},
       year={2022}
   }
   Source: https://github.com/AvivNavon/nash-mtl/blob/main/methods/weight_methods.py


   .. py:method:: weight_grads(grads_forget: List[torch.Tensor], grads_retain: List[torch.Tensor], accelerator) -> torch.Tensor
      :abstractmethod:


      @return scaled_grad



.. py:class:: GradientWeightingMethodNone(/, **data: Any)

   Bases: :py:obj:`GradientWeightingMethod`


   No weighting is applied, takes just the forget gradients

   For debugging/comparison purposes


   .. py:method:: weight_grads(grads_forget: List[torch.Tensor], grads_retain: List[torch.Tensor], accelerator) -> torch.Tensor

      @return scaled_grad



.. py:class:: GradientWeightingMethodSimple(/, **data: Any)

   Bases: :py:obj:`GradientWeightingMethod`


   Fixed weights for each component


   .. py:attribute:: forget_weight
      :type:  float
      :value: 1.0



   .. py:attribute:: retain_weight
      :type:  float
      :value: 1.0



   .. py:method:: weight_grads(grads_forget: List[torch.Tensor], grads_retain: List[torch.Tensor], accelerator) -> torch.Tensor

      @return scaled_grad



.. py:class:: GradientWeightingMethodMunba(/, **data: Any)

   Bases: :py:obj:`GradientWeightingMethod`


   Inspired by @misc{wu2025munbamachineunlearningnash,
       title={MUNBa: Machine Unlearning via Nash Bargaining},
       author={Jing Wu and Mehrtash Harandi},
       year={2025},
       eprint={2411.15537},
       archivePrefix={arXiv},
       primaryClass={cs.CV},
       url={https://arxiv.org/pdf/2411.15537v1},
   }
   The closed-form solution is implemented as described in the V1 of the paper.


   .. py:method:: weight_grads(grads_forget: List[torch.Tensor], grads_retain: List[torch.Tensor], accelerator) -> torch.Tensor

      @return scaled_grad