:py:mod:`neural_compressor.adaptor.tf_utils.quantize_graph.qat.quantize_helper`
===============================================================================

.. py:module:: neural_compressor.adaptor.tf_utils.quantize_graph.qat.quantize_helper

.. autoapi-nested-parse::

   QAT Quantize Helper Class.



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


Functions
~~~~~~~~~

.. autoapisummary::

   neural_compressor.adaptor.tf_utils.quantize_graph.qat.quantize_helper.init_quantize_config
   neural_compressor.adaptor.tf_utils.quantize_graph.qat.quantize_helper.qat_clone_function



.. py:function:: init_quantize_config(model, quantize_recipe=None)

   Initialize quantization config at the beginning of QAT process.

   :param model_name: Special pre-optimized model name.
   :type model_name: string
   :param quantize_recipe: A dict that decide whether given layers should be quantized.
   :type quantize_recipe: dict

   :returns:

             QuantizeConfig instance used to decide whether a specific layer
                                      should be quantized.
   :rtype: config (QuantizeConfig)


.. py:function:: qat_clone_function(layer)

   Wrap or leave given layer based on quantize config object parameters.

   :param layer: input Keras layer
   :type layer: tf.keras.layers.Layer

   :returns: layer wrapped by QuantizeWrapper class.
   :rtype: wrapped_layer (QuantizeWrapper)