objectscope.trainer
===================

.. py:module:: objectscope.trainer


Classes
-------

.. autoapisummary::

   objectscope.trainer.TrainSession


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

.. py:class:: TrainSession(train_img_dir, train_coco_json_file, test_img_dir, test_coco_json_file, config_file_url, num_classes, train_data_name=None, test_data_name=None, train_metadata={}, test_metadata={}, output_dir='output/object_detector', device='cuda', num_workers=12, imgs_per_batch=4, base_lr=5e-05, max_iter=5000, checkpoint_period=50, **kwargs)

   Bases: :py:obj:`object`


   .. py:attribute:: train_img_dir


   .. py:attribute:: train_coco_json_file


   .. py:attribute:: test_img_dir


   .. py:attribute:: test_coco_json_file


   .. py:attribute:: train_metadata


   .. py:attribute:: test_metadata


   .. py:attribute:: config_file_url


   .. py:attribute:: num_classes


   .. py:attribute:: output_dir
      :value: 'output/object_detector'



   .. py:attribute:: device
      :value: 'cuda'



   .. py:attribute:: num_workers
      :value: 12



   .. py:attribute:: imgs_per_batch
      :value: 4



   .. py:attribute:: base_lr
      :value: 5e-05



   .. py:attribute:: max_iter
      :value: 5000



   .. py:attribute:: checkpoint_period
      :value: 50



   .. py:attribute:: test_data_name
      :value: None



   .. py:attribute:: train_data_name
      :value: None



   .. py:method:: register_dataset(train_img_dir=None, train_coco_json_file=None, test_img_dir=None, test_coco_json_file=None, train_data_name=None, test_data_name=None, train_metadata={}, test_metadata={})


   .. py:method:: create_config(num_classes=None, config_file_url=None, num_workers=None, imgs_per_batch=None, base_lr=None, max_iter=None, checkpoint_period=None, output_dir=None, device=None, train_data_name=None, test_data_name=None, anchor_ratios: Union[None, List[List]] = None, anchor_sizes: Union[None, List[List]] = None, evaluate_period=1)

      _summary_

      :param num_classes: _description_. Defaults to None.
      :type num_classes: _type_, optional
      :param config_file_url: _description_. Defaults to None.
      :type config_file_url: _type_, optional
      :param num_workers: _description_. Defaults to None.
      :type num_workers: _type_, optional
      :param imgs_per_batch: _description_. Defaults to None.
      :type imgs_per_batch: _type_, optional
      :param base_lr: _description_. Defaults to None.
      :type base_lr: _type_, optional
      :param max_iter: _description_. Defaults to None.
      :type max_iter: _type_, optional
      :param checkpoint_period: _description_. Defaults to None.
      :type checkpoint_period: _type_, optional
      :param output_dir: _description_. Defaults to None.
      :type output_dir: _type_, optional
      :param device: _description_. Defaults to None.
      :type device: _type_, optional
      :param train_data_name: _description_. Defaults to None.
      :type train_data_name: _type_, optional
      :param test_data_name: _description_. Defaults to None.
      :type test_data_name: _type_, optional
      :param anchor_ratios: Anchor ratios use for generating anchor boxes. Example [[0.7685566328549631, 1.8715268243900367, 1.1942387054643602]].
      :type anchor_ratios: Union[None, List[List]], optional
      :param anchor_sizes: Anchor sizes use for generating anchor boxes and RPN. Example [[240.9236833908755], [59.864835712691715], [153.60699447681742], [434.33823627084996], [103.37411650130916]].
      :type anchor_sizes: list, optional
      :param evaluate_period: _description_. Defaults to 1.
      :type evaluate_period: int, optional

      :returns: _description_
      :rtype: _type_



   .. py:method:: get_trainer(cfg=None)


   .. py:method:: create_trainer()


   .. py:method:: run()