deepmd.tf.nvnmd.utils.config
============================

.. py:module:: deepmd.tf.nvnmd.utils.config


Attributes
----------

.. autoapisummary::

   deepmd.tf.nvnmd.utils.config.log
   deepmd.tf.nvnmd.utils.config.nvnmd_cfg


Classes
-------

.. autoapisummary::

   deepmd.tf.nvnmd.utils.config.NvnmdConfig


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

.. py:data:: log

.. py:class:: NvnmdConfig(jdata: dict)

   
   Configuration for NVNMD
   record the message of model such as size, using nvnmd or not.


   :Parameters:

       **jdata**
           a dictionary of input script












   .. rubric:: References

   DOI: 10.1038/s41524-022-00773-z

   .. only:: latex

      


   ..
       !! processed by numpydoc !!

   .. py:attribute:: version
      :value: 0



   .. py:attribute:: enable
      :value: False



   .. py:attribute:: map


   .. py:attribute:: config


   .. py:attribute:: save_path
      :value: 'nvnmd/config.npy'



   .. py:attribute:: weight


   .. py:method:: init_from_jdata(jdata: dict = {}) -> None

      
      Initialize this class with `jdata` loaded from input script.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_value() -> None

      
      Initialize member with dict.
















      ..
          !! processed by numpydoc !!


   .. py:method:: update_config() -> None

      
      Update config from dict.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_train_mode(mod='cnn') -> None

      
      Configure for taining cnn or qnn.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_from_config(jdata) -> None

      
      Initialize member element one by one.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_config_by_version(version, max_nnei) -> None

      
      Initialize version-dependent parameters.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_net_size() -> None

      
      Initialize net_size.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_from_deepmd_input(jdata) -> None

      
      Initialize members with input script of deepmd.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_dscp(jdata: dict, jdata_parent: dict = {}) -> dict

      
      Initialize members about descriptor.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_fitn(jdata: dict, jdata_parent: dict = {}) -> dict

      
      Initialize members about fitting network.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_dpin(jdata: dict, jdata_parent: dict = {}) -> dict

      
      Initialize members about other deepmd input.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_size(jdata: dict, jdata_parent: dict = {}) -> dict

      
      Initialize members about ram capacity.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_ctrl(jdata: dict, jdata_parent: dict = {}) -> dict

      
      Initialize members about control signal.
















      ..
          !! processed by numpydoc !!


   .. py:method:: init_nbit(jdata: dict, jdata_parent: dict = {}) -> dict

      
      Initialize members about quantification precision.
















      ..
          !! processed by numpydoc !!


   .. py:method:: save(file_name=None) -> None

      
      Save all configuration to file.
















      ..
          !! processed by numpydoc !!


   .. py:method:: set_ntype(ntype) -> None

      
      Set the number of type.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_s_range(davg, dstd) -> None

      
      Get the range of switch function.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_dscp_jdata()

      
      Generate `model/descriptor` in input script.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_fitn_jdata()

      
      Generate `model/fitting_net` in input script.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_model_jdata()

      
      Generate `model` in input script.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_nvnmd_jdata()

      
      Generate `nvnmd` in input script.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_learning_rate_jdata()

      
      Generate `learning_rate` in input script.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_loss_jdata()

      
      Generate `loss` in input script.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_training_jdata()

      
      Generate `training` in input script.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_deepmd_jdata()

      
      Generate input script with member element one by one.
















      ..
          !! processed by numpydoc !!


   .. py:method:: get_dp_init_weights()

      
      Build the weight dict for initialization of net.
















      ..
          !! processed by numpydoc !!


   .. py:method:: disp_message() -> None

      
      Display the log of NVNMD.
















      ..
          !! processed by numpydoc !!


.. py:data:: nvnmd_cfg