dacapo.experiments.run
======================

.. py:module:: dacapo.experiments.run


Classes
-------

.. autoapisummary::

   dacapo.experiments.run.Run


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

.. py:class:: Run(run_config, load_starter_model: bool = True)

   Class representing a run in the experiment. A run is a combination of a task, architecture, trainer, datasplit,
   model, optimizer, training stats, and validation scores. It also contains the name of the run, the number of
   iterations to train for, and the interval at which to validate. It also contains a start object that can be used to
   initialize the model with preloaded weights. The run object can be used to move the optimizer to a specified device.

   .. attribute:: name

      The name of the run.

      :type: str

   .. attribute:: train_until

      The number of iterations to train for.

      :type: int

   .. attribute:: validation_interval

      The interval at which to validate.

      :type: int

   .. attribute:: task

      The task object.

      :type: Task

   .. attribute:: architecture

      The architecture object.

      :type: Architecture

   .. attribute:: trainer

      The trainer object.

      :type: Trainer

   .. attribute:: datasplit

      The datasplit object.

      :type: DataSplit

   .. attribute:: model

      The model object.

      :type: Model

   .. attribute:: optimizer

      The optimizer object.

      :type: torch.optim.Optimizer

   .. attribute:: training_stats

      The training stats object.

      :type: TrainingStats

   .. attribute:: validation_scores

      The validation scores object.

      :type: ValidationScores

   .. attribute:: start

      The start object.

      :type: Start

   .. method:: move_optimizer(device

      torch.device, empty_cuda_cache: bool) -> None:
      Moves the optimizer to the specified device.

   .. method:: get_validation_scores(run_config) -> ValidationScores

      
      Static method to get the validation scores without initializing model, optimizer, trainer, etc.

   .. note::

      The iteration stats list is structured as follows:
      - The outer list contains the stats for each iteration.
      - The inner list contains the stats for each training iteration.


   .. py:attribute:: name
      :type:  str


   .. py:attribute:: train_until
      :type:  int


   .. py:attribute:: validation_interval
      :type:  int


   .. py:attribute:: task
      :type:  dacapo.experiments.tasks.task.Task


   .. py:attribute:: architecture
      :type:  dacapo.experiments.architectures.Architecture


   .. py:attribute:: trainer
      :type:  dacapo.experiments.trainers.Trainer


   .. py:attribute:: model
      :type:  dacapo.experiments.model.Model


   .. py:attribute:: optimizer
      :type:  torch.optim.Optimizer


   .. py:attribute:: training_stats
      :type:  dacapo.experiments.training_stats.TrainingStats


   .. py:attribute:: task_type


   .. py:attribute:: architecture_type


   .. py:attribute:: trainer_type


   .. py:attribute:: datasplit_type


   .. py:attribute:: start


   .. py:attribute:: new_head
      :value: None



   .. py:property:: datasplit


   .. py:property:: validation_scores


   .. py:method:: get_validation_scores(run_config) -> dacapo.experiments.validation_scores.ValidationScores
      :staticmethod:


      Static method to get the validation scores without initializing model, optimizer, trainer, etc.

      :param run_config: The configuration for the run.

      :returns: The validation scores.

      :raises AssertionError: If the task or datasplit types are not specified in the run_config.

      .. rubric:: Examples

      >>> validation_scores = Run.get_validation_scores(run_config)
      >>> validation_scores
      ValidationScores object



   .. py:method:: move_optimizer(device: torch.device, empty_cuda_cache: bool = False) -> None

      Moves the optimizer to the specified device.

      :param device: The device to move the optimizer to.
      :param empty_cuda_cache: Whether to empty the CUDA cache after moving the optimizer.

      :raises AssertionError: If the optimizer state is not a dictionary.

      .. rubric:: Examples

      >>> run.move_optimizer(device)
      >>> run.optimizer
      Optimizer object



   .. py:method:: visualize_pipeline(bind_address='0.0.0.0', bind_port=0)

      Visualizes the pipeline for the run, including all produced arrays.

      :param bind_address: str
                           Bind address for Neuroglancer webserver
      :param bind_port: int
                        Bind port for Neuroglancer webserver

      .. rubric:: Examples

      >>> run.visualize_pipeline()