8442321b9b7a7cef7e3806b525965c36cea1549e,examples/60_search/example_parallel_manual_spawning.py,,,#,97

Before Change

// Build and fit the classifier
// ============================

processes = []
spawn_classifier = get_spawn_classifier(X_train, y_train)
for i in range(4):  // set this at roughly half of your cores
    p = multiprocessing.Process(
        target=spawn_classifier,
        args=(i, "breast_cancer"),
    )
    p.start()
    processes.append(p)
for p in processes:
    p.join()

print("Starting to build an ensemble!")
automl = AutoSklearnClassifier(
    time_left_for_this_task=30,
    per_run_time_limit=15,

After Change

            // Pass a dask client which connects to the previously constructed cluster.
            dask_client=client,
        )
        automl.fit(X_train, y_train)

        automl.fit_ensemble(
            y_train,
            task=MULTICLASS_CLASSIFICATION,

Instances