TQDMProgressBar

class pytorch_lightning.callbacks.TQDMProgressBar(refresh_rate=1, process_position=0)

Bases: pytorch_lightning.callbacks.progress.base.ProgressBarBase

This is the default progress bar used by Lightning. It prints to stdout using the tqdm package and shows up to four different bars:

• sanity check progress: the progress during the sanity check run

• main progress: shows training + validation progress combined. It also accounts for multiple validation runs during training when val_check_interval is used.

• validation progress: only visible during validation; shows total progress over all validation datasets.

• test progress: only active when testing; shows total progress over all test datasets.

For infinite datasets, the progress bar never ends.

If you want to customize the default tqdm progress bars used by Lightning, you can override specific methods of the callback class and pass your custom implementation to the Trainer.

Example

>>> class LitProgressBar(TQDMProgressBar):
...     def init_validation_tqdm(self):
...         bar = super().init_validation_tqdm()
...         bar.set_description('running validation ...')
...         return bar
...
>>> bar = LitProgressBar()
>>> from pytorch_lightning import Trainer
>>> trainer = Trainer(callbacks=[bar])

Parameters:

• refresh_rate (int) – Determines at which rate (in number of batches) the progress bars get updated. Set it to 0 to disable the display.

• process_position (int) – Set this to a value greater than 0 to offset the progress bars by this many lines. This is useful when you have progress bars defined elsewhere and want to show all of them together. This corresponds to process_position in the Trainer.

disable()

You should provide a way to disable the progress bar.

Return type:

None

enable()

You should provide a way to enable the progress bar.

The Trainer will call this in e.g. pre-training routines like the learning rate finder. to temporarily enable and disable the main progress bar.

Return type:

None

init_predict_tqdm()

Override this to customize the tqdm bar for predicting.

Return type:

Tqdm

init_sanity_tqdm()

Override this to customize the tqdm bar for the validation sanity run.

Return type:

Tqdm

init_test_tqdm()

Override this to customize the tqdm bar for testing.

Return type:

Tqdm

init_train_tqdm()

Override this to customize the tqdm bar for training.

Return type:

Tqdm

init_validation_tqdm()

Override this to customize the tqdm bar for validation.

Return type:

Tqdm

on_predict_batch_end(*_)

Called when the predict batch ends.

Return type:

None

on_predict_batch_start(trainer, pl_module, batch, batch_idx, dataloader_idx)

Called when the predict batch begins.

Return type:

None

on_predict_end(trainer, pl_module)

Called when predict ends.

Return type:

None

on_predict_start(trainer, pl_module)

Called when the predict begins.

Return type:

None

on_sanity_check_end(*_)

Called when the validation sanity check ends.

Return type:

None

on_sanity_check_start(*_)

Called when the validation sanity check starts.

Return type:

None

on_test_batch_end(*_)

Called when the test batch ends.

Return type:

None

on_test_batch_start(trainer, pl_module, batch, batch_idx, dataloader_idx)

Called when the test batch begins.

Return type:

None

on_test_end(trainer, pl_module)

Called when the test ends.

Return type:

None

on_test_start(trainer, pl_module)

Called when the test begins.

Return type:

None

on_train_batch_end(trainer, pl_module, *_)

Called when the train batch ends. :rtype: None

Note

The value outputs["loss"] here will be the normalized value w.r.t accumulate_grad_batches of the loss returned from training_step.

on_train_end(*_)

Called when the train ends.

Return type:

None

on_train_epoch_end(trainer, pl_module)

Called when the train epoch ends.

To access all batch outputs at the end of the epoch, either: :rtype: None

1. Implement training_epoch_end in the LightningModule and access outputs via the module OR

2. Cache data across train batch hooks inside the callback implementation to post-process in this hook.

on_train_epoch_start(trainer, *_)

Called when the train epoch begins.

Return type:

None

on_train_start(*_)

Called when the train begins.

Return type:

None

on_validation_batch_end(trainer, *_)

Called when the validation batch ends.

Return type:

None

on_validation_batch_start(trainer, pl_module, batch, batch_idx, dataloader_idx)

Called when the validation batch begins.

Return type:

None

on_validation_end(trainer, pl_module)

Called when the validation loop ends.

Return type:

None

on_validation_start(trainer, pl_module)

Called when the validation loop begins.

Return type:

None

print(*args, sep=' ', **kwargs)

You should provide a way to print without breaking the progress bar.

Return type:

None