CometCallback

class radionets.core.callbacks.CometCallback(name, validation_data, plot_n_epochs, amp_phase, scale)

Bases: Callback

Callback for logging training metrics and visualizations to Comet ML.

This callback logs training and validation losses, and creates plots for predictions and Fourier-transformed data for monitoring during training.

Parameters:

namestr

Project name for the Comet ML experiment.

validation_datastr or Path

Path to the validation dataset.

plot_n_epochsint

Frequency of plotting (every n epochs).

amp_phasebool

Whether to use amplitude-phase representation.

scalestr

Scaling method for data.

Attributes Summary

learn
name	Name of the Callback, camel-cased and with 'Callback' removed
order
run
run_train
run_valid

Methods Summary

__call__(event_name)	Call self.{event_name} if it's defined
after_epoch()
after_train()
after_validate()
plot_test_fft()
plot_test_pred()

Attributes Documentation

learn = None

name

Name of the Callback, camel-cased and with ‘Callback’ removed

order = 0

run = True

run_train = True

run_valid = True

Methods Documentation

__call__(event_name)

Call self.{event_name} if it’s defined

after_epoch()

after_train()

after_validate()

plot_test_fft()

plot_test_pred()