:py:mod:`neural_compressor.compression.pruner.model_slim.weight_slim`
=====================================================================

.. py:module:: neural_compressor.compression.pruner.model_slim.weight_slim

.. autoapi-nested-parse::

   Weight Squeezer.



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

Classes
~~~~~~~

.. autoapisummary::

   neural_compressor.compression.pruner.model_slim.weight_slim.PostCompressionUtils
   neural_compressor.compression.pruner.model_slim.weight_slim.LinearCompression
   neural_compressor.compression.pruner.model_slim.weight_slim.LinearCompressionIterator




.. py:class:: PostCompressionUtils




   Operations library related to weight compression.


.. py:class:: LinearCompression(root_linear, target_linears)




   Class which automatically compresses two consecutive linear layers.

   For two consecutive linear layer, when the second layer's input channel is pruned,
   then the first layer's output channel can also be pruned,
   while the second layer's output hidden state value is identical.
   for example, two consecutive linears have following structure.
   x = layer_1(input)
   x = act_fn(x)
   x = layer_2(x)

   :param layer_1: the first Linear layer.
   :type layer_1: torch.nn.Linear
   :param layer_2: the second Linear layer.
   :type layer_2: torch.nn.Linear

   .. attribute:: layer_1

      the first Linear layer.

      :type: torch.nn.Linear

   .. attribute:: layer_2

      the second Linear layer.

      :type: torch.nn.Linear

   .. attribute:: device

      the device of layers' weights.


.. py:class:: LinearCompressionIterator(linear_patterns)




   Pruner of a sequence of consecutive linear patterns.

   .. attribute:: linear_patterns

      a iterable object of consecutive linear patterns.

      :type: dict/list