:orphan:

:py:mod:`neural_compressor.torch.algorithms.weight_only.teq`
============================================================

.. py:module:: neural_compressor.torch.algorithms.weight_only.teq


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

Classes
~~~~~~~

.. autoapisummary::

   neural_compressor.torch.algorithms.weight_only.teq.TrainableEquivalentTransformation
   neural_compressor.torch.algorithms.weight_only.teq.TEQuantizer




.. py:class:: TrainableEquivalentTransformation(model, weight_config={}, absorb_to_layer={}, folding=True, example_inputs=None)


   Weight-only quantization, Trainable Equivalent Transformation (TEQ).


.. py:class:: TEQuantizer(quant_config, folding, absorb_to_layer, example_inputs)




   The base quantizer for all algorithm quantizers.

   The `Quantizer` unifies the interfaces across various quantization algorithms, including GPTQ, RTN, etc.
   Given a float model, `Quantizer` apply the quantization algorithm to the model according to the `quant_config`.

   To implement a new quantization algorithm,, inherit from `Quantizer` and implement the following methods:
       - `prepare`: prepare a given model for convert.
       - `convert`: convert a prepared model to a quantized model.
   Note: `quantize` and `execute` are optional for new quantization algorithms.