:py:mod:`medcat.utils.checkpoint`
=================================

.. py:module:: medcat.utils.checkpoint


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

Classes
~~~~~~~

.. autoapisummary::

   medcat.utils.checkpoint.Checkpoint
   medcat.utils.checkpoint.CheckpointConfig
   medcat.utils.checkpoint.CheckpointManager




Attributes
~~~~~~~~~~

.. autoapisummary::

   medcat.utils.checkpoint.T
   medcat.utils.checkpoint.logger


.. py:data:: T

   

.. py:data:: logger

   

.. py:class:: Checkpoint(dir_path, *, steps = DEFAULT_STEP, max_to_keep = DEFAULT_MAX_TO_KEEP)


   Bases: :py:obj:`object`

   The base class of checkpoint objects

   :param dir_path: The path to the parent directory of checkpoint files.
   :type dir_path: str
   :param steps: The number of processed sentences/documents before a checkpoint is saved
                 (N.B.: A small number could result in error "no space left on device"),
   :type steps: int
   :param max_to_keep: The maximum number of checkpoints to keep
                       (N.B.: A large number could result in error "no space left on device").
   :type max_to_keep: int

   .. py:property:: steps
      :type: int


   .. py:property:: max_to_keep
      :type: int


   .. py:property:: count
      :type: int


   .. py:property:: dir_path
      :type: str


   .. py:attribute:: DEFAULT_STEP
      :value: 1000

      

   .. py:attribute:: DEFAULT_MAX_TO_KEEP
      :value: 1

      

   .. py:method:: __init__(dir_path, *, steps = DEFAULT_STEP, max_to_keep = DEFAULT_MAX_TO_KEEP)


   .. py:method:: from_latest(dir_path)
      :classmethod:

      Retrieve the latest checkpoint from the parent directory.

      :param dir_path: The path to the directory containing checkpoint files.
      :type dir_path: str

      :Returns: **T** -- A new checkpoint object.

      :raises Exception: If no checkpoint is found.


   .. py:method:: save(cdb, count)

      Save the CDB as the latest checkpoint.

      :param cdb: The MedCAT CDB object to be checkpointed.
      :type cdb: CDB
      :param count: The number of the finished steps.
      :type count: int


   .. py:method:: restore_latest_cdb()

      Restore the CDB from the latest checkpoint.

      :Returns: **cdb** (*CDB*) -- The MedCAT CDB object.

      :raises Exception: If no checkpoint is found.


   .. py:method:: _get_ckpt_file_paths(dir_path)
      :staticmethod:


   .. py:method:: _get_steps_and_count(file_path)
      :staticmethod:



.. py:class:: CheckpointConfig


   Bases: :py:obj:`object`

   .. py:attribute:: output_dir
      :type: str
      :value: 'checkpoints'

      

   .. py:attribute:: steps
      :type: int

      

   .. py:attribute:: max_to_keep
      :type: int

      


.. py:class:: CheckpointManager(name, checkpoint_config)


   Bases: :py:obj:`object`

   The class for managing checkpoints of specific training type and their configuration

   :param name: The name of the checkpoint manager (also used as the checkpoint base directory name).
   :type name: str
   :param checkpoint_config: The checkpoint config object.
   :type checkpoint_config: medcat.utils.checkpoint.CheckpointConfig

   .. py:method:: __init__(name, checkpoint_config)


   .. py:method:: create_checkpoint(dir_path = None)

      Create a new checkpoint inside the checkpoint base directory.

      :param dir_path: The path to the checkpoint directory.
      :type dir_path: str

      :Returns: **CheckPoint** -- A checkpoint object.


   .. py:method:: get_latest_checkpoint(base_dir_path = None)

      Retrieve the latest checkpoint from the checkpoint base directory.

      :param base_dir_path: The path to the directory containing checkpoint files.
      :type base_dir_path: string

      :Returns: **CheckPoint** -- A checkpoint object


   .. py:method:: get_latest_training_dir(base_dir_path)
      :classmethod:

      Retrieve the latest training directory containing all checkpoints.

      :param base_dir_path: The path to the directory containing all checkpointed trainings.
      :type base_dir_path: str

      :Returns: **str** -- The path to the latest training directory containing all checkpoints.

      :raises ValueError: If no checkpoint is found.