Callback for Study.optimize

This tutorial showcases how to use & implement Optuna Callback for optimize().

Callback is called after every evaluation of objective, and it takes Study and FrozenTrial as arguments, and does some work.

MLflowCallback is a great example.

Stop optimization after some trials are pruned in a row

This example implements a stateful callback which stops the optimization if a certain number of trials are pruned in a row. The number of trials pruned in a row is specified by threshold.

import optuna

class StopWhenTrialKeepBeingPrunedCallback:
    def __init__(self, threshold: int):
        self.threshold = threshold
        self._consequtive_pruned_count = 0

    def __call__(self, study:, trial: optuna.trial.FrozenTrial) -> None:
        if trial.state == optuna.trial.TrialState.PRUNED:
            self._consequtive_pruned_count += 1
            self._consequtive_pruned_count = 0

        if self._consequtive_pruned_count >= self.threshold:

This objective prunes all the trials except for the first 5 trials (trial.number starts with 0).

def objective(trial):
    if trial.number > 4:
        raise optuna.TrialPruned

    return trial.suggest_float("x", 0, 1)

Here, we set the threshold to 2: optimization finishes once two trials are pruned in a row. So, we expect this study to stop after 7 trials.

import logging
import sys

# Add stream handler of stdout to show the messages

study_stop_cb = StopWhenTrialKeepBeingPrunedCallback(2)
study = optuna.create_study()
study.optimize(objective, n_trials=10, callbacks=[study_stop_cb])
A new study created in memory with name: no-name-59b4ca7d-6029-4c33-b925-a186fc78dfad
Trial 0 finished with value: 0.6763855742608957 and parameters: {'x': 0.6763855742608957}. Best is trial 0 with value: 0.6763855742608957.
Trial 1 finished with value: 0.35150210997232645 and parameters: {'x': 0.35150210997232645}. Best is trial 1 with value: 0.35150210997232645.
Trial 2 finished with value: 0.2030024738647338 and parameters: {'x': 0.2030024738647338}. Best is trial 2 with value: 0.2030024738647338.
Trial 3 finished with value: 0.9513004149158926 and parameters: {'x': 0.9513004149158926}. Best is trial 2 with value: 0.2030024738647338.
Trial 4 finished with value: 0.9462275958606968 and parameters: {'x': 0.9462275958606968}. Best is trial 2 with value: 0.2030024738647338.
Trial 5 pruned.
Trial 6 pruned.

As you can see in the log above, the study stopped after 7 trials as expected.

Total running time of the script: (0 minutes 0.005 seconds)

Gallery generated by Sphinx-Gallery