:py:mod:`neural_compressor.adaptor.torch_utils.bf16_convert`
============================================================

.. py:module:: neural_compressor.adaptor.torch_utils.bf16_convert

.. autoapi-nested-parse::

   Bf16 Convert for Torch Utils.



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

Classes
~~~~~~~

.. autoapisummary::

   neural_compressor.adaptor.torch_utils.bf16_convert.BF16ModuleWrapper



Functions
~~~~~~~~~

.. autoapisummary::

   neural_compressor.adaptor.torch_utils.bf16_convert.Convert
   neural_compressor.adaptor.torch_utils.bf16_convert.bf16_symbolic_trace



.. py:class:: BF16ModuleWrapper(module)




   BF16Module Wrapper Class.


.. py:function:: Convert(model, tune_cfg)

   Convert to bf16 model.

   :param model: the input model.
   :type model: object
   :param tune_cfg: dictionary of quantization configuration.
   :type tune_cfg: dict

   :returns: model with mixed precision.
   :rtype: mixed_precision_model (object)


.. py:function:: bf16_symbolic_trace(model, fx_sub_module_list, prefix='')

   Symbolic trace for bf16 models.

   :param model: the input model.
   :type model: object
   :param fx_sub_module_list: _description_
   :type fx_sub_module_list: list
   :param prefix: prefix of op name.
   :type prefix: str

   :returns: model (object)