TQDMProgressBar

class braintools.trainer.TQDMProgressBar(refresh_rate=1, process_position=0, leave=True)

TQDM progress bar for training visualization.

Requires the ‘tqdm’ package to be installed.

Parameters:

• refresh_rate (int) – Number of batches between progress bar updates.

• process_position (int) – Position of the progress bar.

• leave (bool) – Whether to leave the progress bar after completion.

Examples

>>> progress = TQDMProgressBar()
>>> trainer = Trainer(callbacks=[progress], enable_progress_bar=False)

on_train_batch_end(trainer, module, outputs, batch, batch_idx)

Update progress bar.

on_train_epoch_end(trainer, module)

Close progress bar.

on_train_epoch_start(trainer, module)

Create progress bar for epoch.