avalanche.evaluation.metrics.ExperienceForgetting

class avalanche.evaluation.metrics.ExperienceForgetting[source]

The ExperienceForgetting metric, describing the accuracy loss detected for a certain experience.

This plugin metric, computed separately for each experience, is the difference between the accuracy result obtained after first training on a experience and the accuracy result obtained on the same experience at the end of successive experiences.

This metric is computed during the eval phase only.

__init__()[source]

Creates an instance of the ExperienceForgetting metric.

Methods

__init__()

Creates an instance of the ExperienceForgetting metric.

after_backward(strategy)

after_eval(strategy)

after_eval_dataset_adaptation(strategy)

after_eval_exp(strategy)

after_eval_forward(strategy)

after_eval_iteration(strategy)

after_forward(strategy)

after_train_dataset_adaptation(strategy)

after_training(strategy)

after_training_epoch(strategy)

after_training_exp(strategy)

after_training_iteration(strategy)

after_update(strategy)

before_backward(strategy)

before_eval(strategy)

before_eval_dataset_adaptation(strategy)

before_eval_exp(strategy)

before_eval_forward(strategy)

before_eval_iteration(strategy)

before_forward(strategy)

before_train_dataset_adaptation(strategy)

before_training(strategy)

before_training_epoch(strategy)

before_training_exp(strategy)

before_training_iteration(strategy)

before_update(strategy)

metric_result(strategy)

metric_update(strategy)

reset()

Resets the metric.

reset_last()

Resets the last metric value.

result()

Forgetting for all experiences.

result_key(k)

Forgetting for an experience defined by its key.

update(k, v[, initial])

Update forgetting metric.

Attributes

forgetting

The general metric to compute forgetting

eval_exp_id

The current evaluation experience id

train_exp_id

The last encountered training experience id