base_model¶
Base model implementing helper methods.
Module Contents¶
-
class
base_model.BaseModel(hparams)¶ Bases:
pytorch_lightning.core.LightningModule
The primary class containing all the training functionality. It is equivalent toPyTorch nn.Module in all aspects.
- Parameters
LightningModule (nn.Module) – The Pytorch-Lightning module derived from nn.module withuseful hooks
- Raises
NotImplementedError – Some methods must be overridden
-
abstract
forward(self)¶ Dummy method to do forward pass on the model.
- Raises
NotImplementedError – The method must be overridden in the derived models
-
training_step(self, batch, batch_idx)¶ Called inside the testing loop with the data from the testing dataloader passed in as batch. The implementation is delegated to the dataloader instead.
For performance critical usecase prefer monkey-patching instead.
- Parameters
model (Model) – The chosen model
batch (int) – Batch of input and ground truth variables
- Returns
Loss and logs
- Return type
dict
-
validation_step(self, batch, batch_idx)¶ Called inside the validation loop with the data from the validation dataloader passed in as batch. The implementation is delegated to the dataloader instead.
For performance critical usecase prefer monkey-patching instead.
- Parameters
model (Model) – The chosen model
batch (int) – Batch of input and ground truth variables
- Returns
Loss and logs
- Return type
dict
-
test_step(self, batch, batch_idx)¶ Called inside the testing loop with the data from the testing dataloader passed in as batch. The implementation is delegated to the dataloader instead.
For performance critical usecase prefer monkey-patching instead.
- Parameters
model (Model) – The chosen model
batch (int) – Batch of input and ground truth variables
- Returns
Loss and logs
- Return type
dict
-
training_epoch_end(self, outputs)¶ Called at the end of training epoch to aggregate outputs.
- Parameters
outputs (list) – List of individual outputs of each training step.
- Returns
Loss and logs.
- Return type
dict
-
validation_epoch_end(self, outputs)¶ Called at the end of validation epoch to aggregate outputs.
- Parameters
outputs (list) – List of individual outputs of each validation step.
- Returns
Loss and logs.
- Return type
dict
-
test_epoch_end(self, outputs)¶ Called at the end of testing epoch to aggregate outputs.
- Parameters
outputs (list) – List of individual outputs of each testing step.
- Returns
Loss and logs.
- Return type
dict
-
configure_optimizers(self)¶ Decide optimizers and learning rate schedulers.
At least one optimizer is required.
- Returns
Optimizer and the schedular
- Return type
tuple
-
add_bias(self, bias)¶ Initialize bias parameter of the last layer with the output variable’s mean.
- Parameters
bias (float) – Mean of the output variable.
-
prepare_data(self, ModelDataset=None, force=False)¶ Load and split the data for training and test during the first call. Behavior on second call determined by the force parameter.
- Parameters
ModelDataset (class, optional) – The dataset class to be used with the model, defaults to None
force (bool, optional) – Force the data preperation even if already prepared, defaults to False
-
train_dataloader(self)¶ Create the training dataloader from the training dataset.
- Returns
The training dataloader
- Return type
Dataloader
-
val_dataloader(self)¶ Create the validation dataloader from the validation dataset.
- Returns
The validation dataloader
- Return type
Dataloader
-
test_dataloader(self)¶ Create the testing dataloader from the testing dataset.
- Returns
The testing dataloader
- Return type
Dataloader