:py:mod:`neural_compressor.experimental.nas.dynas` ================================================== .. py:module:: neural_compressor.experimental.nas.dynas .. autoapi-nested-parse:: DyNAS approach class. Module Contents --------------- Classes ~~~~~~~ .. autoapisummary:: neural_compressor.experimental.nas.dynas.DyNAS .. py:class:: DyNAS(conf_fname_or_obj) Bases: :py:obj:`neural_compressor.experimental.nas.nas.NASBase` DyNAS approach. Defining the pipeline for DyNAS approach. :param conf_fname_or_obj: The path to the YAML configuration file or the object of NASConfig. :type conf_fname_or_obj: string or obj .. py:method:: estimate(individual) Estimate performance of the model. :returns: Evaluated metrics of the model. .. py:method:: init_for_search() Initialize the search configuration. .. py:method:: search() Execute the search process. :returns: Best model architectures found in the search process. .. py:method:: select_model_arch() Select the model architecture. .. py:method:: create_acc_predictor() Create the accuracy predictor. .. py:method:: create_macs_predictor() Create the MACs predictor. .. py:method:: create_latency_predictor() Create the latency predictor. .. py:method:: init_cfg(conf_fname_or_obj) Initialize the configuration.