ThresholdStopper#
- class composer.callbacks.ThresholdStopper(monitor, dataloader_label, threshold, *, comp=None, stop_on_batch=False)[source]#
Halt training when a metric value reaches a certain threshold.
Example
>>> from composer import Evaluator, Trainer >>> from composer.callbacks.threshold_stopper import ThresholdStopper >>> # constructing trainer object with this callback >>> threshold_stopper = ThresholdStopper('MulticlassAccuracy', 'my_evaluator', 0.7) >>> evaluator = Evaluator( ... dataloader = eval_dataloader, ... label = 'my_evaluator', ... metric_names = ['MulticlassAccuracy'] ... ) >>> trainer = Trainer( ... model=model, ... train_dataloader=train_dataloader, ... eval_dataloader=evaluator, ... optimizers=optimizer, ... max_duration="1ep", ... callbacks=[threshold_stopper], ... )
- Parameters
monitor (str) โ The name of the metric to monitor.
dataloader_label (str) โ The label of the dataloader or evaluator associated with the tracked metric. If monitor is in an Evaluator, the dataloader_label field should be set to the Evaluatorโs label. If monitor is a training metric or an ordinary evaluation metric not in an Evaluator, dataloader_label should be set to โtrainโ or โevalโ respectively. If dataloader_label is set to โtrainโ, then the callback will stop training in the middle of the epoch.
threshold (float) โ The threshold that dictates when to halt training. Whether training stops if the metric exceeds or falls below the threshold depends on the comparison operator.
comp (Callable[[Any, Any], Any], optional) โ A comparison operator to measure change of the monitored metric. The comparison operator will be called
comp(current_value, prev_best)
. For metrics where the optimal value is low (error, loss, perplexity), use a less than operator and for metrics like accuracy where the optimal value is higher, use a greater than operator. Defaults totorch.less()
if loss, error, or perplexity are substrings of the monitored metric, otherwise defaults totorch.greater()
stop_on_batch (bool, optional) โ A bool that indicates whether to stop training in the middle of an epoch if the training metrics satisfy the threshold comparison. Defaults to False.