dicee.trainer.torch_trainer_ddp
===============================

.. py:module:: dicee.trainer.torch_trainer_ddp


Classes
-------

.. autoapisummary::

   dicee.trainer.torch_trainer_ddp.TorchDDPTrainer
   dicee.trainer.torch_trainer_ddp.NodeTrainer


Functions
---------

.. autoapisummary::

   dicee.trainer.torch_trainer_ddp.make_iterable_verbose


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

.. py:function:: make_iterable_verbose(iterable_object, verbose, desc='Default', position=None, leave=True) -> Iterable

.. py:class:: TorchDDPTrainer(args, callbacks)

   Bases: :py:obj:`dicee.abstracts.AbstractTrainer`


    A Trainer based on torch.nn.parallel.DistributedDataParallel

    Arguments
   ----------
   train_set_idx
       Indexed triples for the training.
   entity_idxs
       mapping.
   relation_idxs
       mapping.
   form
       ?
   store
        ?
   label_smoothing_rate
        Using hard targets (0,1) drives weights to infinity.
        An outlier produces enormous gradients.

   :rtype: torch.utils.data.Dataset


   .. py:method:: fit(*args, **kwargs)

      Train model



.. py:class:: NodeTrainer(trainer, model: torch.nn.Module, train_dataset_loader: torch.utils.data.DataLoader, callbacks, num_epochs: int)

   .. py:attribute:: trainer


   .. py:attribute:: local_rank


   .. py:attribute:: global_rank


   .. py:attribute:: optimizer


   .. py:attribute:: train_dataset_loader


   .. py:attribute:: loss_func


   .. py:attribute:: callbacks


   .. py:attribute:: model


   .. py:attribute:: num_epochs


   .. py:attribute:: loss_history
      :value: []



   .. py:attribute:: ctx


   .. py:attribute:: scaler


   .. py:method:: extract_input_outputs(z: list)


   .. py:method:: train()

      Training loop for DDP