Shortcuts

IterationTimeLogger

class torchtnt.framework.callbacks.IterationTimeLogger(logger: Union[TensorBoardLogger, SummaryWriter], moving_avg_window: int = 1, log_every_n_steps: int = 1)

A callback which logs iteration times as scalars to TensorBoard.

Parameters:
  • logger – Either a torchtnt.loggers.tensorboard.TensorBoardLogger or a torch.utils.tensorboard.SummaryWriter instance.
  • moving_avg_window – an optional int to control the moving average window
  • log_every_n_steps – an optional int to control the log frequency
on_train_step_end(state: State, unit: TrainUnit[TTrainData]) None

Hook called after a train step ends.

Docs

Access comprehensive developer documentation for PyTorch

View Docs

Tutorials

Get in-depth tutorials for beginners and advanced developers

View Tutorials

Resources

Find development resources and get your questions answered

View Resources