Command-Line Interface




Create a new trial and suggest parameters.


Show the best trial.


Show a list of trials located at the Pareto front.


Create a new study.


Delete a specified study.

storage upgrade

Upgrade the schema of a storage.


Show a list of studies.

study set-user-attr

Set a user attribute to a study.


Finish a trial, which was created by the ask command.


Show a list of trials.

Optuna provides command-line interface as shown in the above table.

Let us assume you are not in IPython shell and writing Python script files instead. It is totally fine to write scripts like the following:

import optuna

def objective(trial):
    x = trial.suggest_float("x", -10, 10)
    return (x - 2) ** 2

if __name__ == "__main__":
    study = optuna.create_study()
    study.optimize(objective, n_trials=100)
    print("Best value: {} (params: {})\n".format(study.best_value, study.best_params))
Best value: 1.801468914658488e-05 (params: {'x': 1.9957556285333886})

However, if we cannot write objective explicitly in Python code such as developing a new drug in a lab, an interactive way is suitable. In Optuna CLI, Ask-and-Tell Interface style commands provide such an interactive and flexible interface.

Let us assume we minimize the objective value depending on a parameter x in \([-10, 10]\) and objective value is calculated via some experiments by hand. Even so, we can invoke the optimization as follows. Don’t care about --storage sqlite:///example.db for now, which is described in Saving/Resuming Study with RDB Backend.

$ STUDY_NAME=`optuna create-study --storage sqlite:///example.db`
$ optuna ask --storage sqlite:///example.db --study-name $STUDY_NAME --sampler TPESampler \
     --search-space '{"x": {"name": "FloatDistribution", "attributes": {"step": null, "low": -10.0, "high": 10.0, "log": false}}}'

[I 2022-08-20 06:08:53,158] Asked trial 0 with parameters {'x': 2.512238141966016}.
{"number": 0, "params": {"x": 2.512238141966016}}

The argument of --search-space option can be generated by using optuna.distributions.distribution_to_json(), for example, optuna.distributions.distribution_to_json(optuna.distributions.FloatDistribution(-10, 10)). Please refer to optuna.distributions.FloatDistribution and optuna.distributions.IntDistribution for detailed explanations of their arguments.

After conducting an experiment using the suggested parameter in the lab, we store the result to Optuna’s study as follows:

$ optuna tell --storage sqlite:///example.db --study-name $STUDY_NAME --trial-number 0 --values 0.7 --state complete
[I 2022-08-20 06:22:50,888] Told trial 0 with values [0.7] and state TrialState.COMPLETE.

Total running time of the script: (0 minutes 0.436 seconds)

Gallery generated by Sphinx-Gallery