deprecation: Is `frequency` key necessary in `lr_scheduler_config`?

[adosar]

Outline & Motivation

The lr_scheduler_config contains a frequency key as shown below:

lr_scheduler_config = {
    # REQUIRED: The scheduler instance
    "scheduler": lr_scheduler,
    # The unit of the scheduler's step size, could also be 'step'.
    # 'epoch' updates the scheduler on epoch end whereas 'step'
    # updates it after a optimizer update.
    "interval": "epoch",
    # How many epochs/steps should pass between calls to
    # `scheduler.step()`. 1 corresponds to updating the learning
    # rate after every epoch/step.
    "frequency": 1,
    # Metric to monitor for schedulers like `ReduceLROnPlateau`
    "monitor": "val_loss",
    # If set to `True`, will enforce that the value specified 'monitor'
    # is available when the scheduler is updated, thus stopping
    # training if not found. If set to `False`, it will only produce a warning
    "strict": True,
    # If using the `LearningRateMonitor` callback to monitor the
    # learning rate progress, this keyword can be used to specify
    # a custom logged name
    "name": None,
}

The frequency of the learning rate decay is already handled by the torch.optim.lr_scheduler's. For example, in StepLR this is handled by the step_size parameter as following:

# Assuming optimizer uses lr = 0.05 for all groups
# lr = 0.05     if epoch < 30
# lr = 0.005    if 30 <= epoch < 60
# lr = 0.0005   if 60 <= epoch < 90
# ...
scheduler = StepLR(optimizer, step_size=30, gamma=0.1)
for epoch in range(100):
    train(...)
    validate(...)
    scheduler.step()

meaning that {..., 'frequency': 2, ...} and step_size=2 is equivalent to {..., 'frequency': 1, ...} and step_size=4.

Is there any case where the frequency key would be beneficial?

Pitch

No response

Additional context

No response

cc @lantiga @justusschock