All Products
Search

This Product

    Platform For AI:How AutoML works

    Document Center

    Platform For AI:How AutoML works

    Last Updated:Mar 11, 2026

    AutoML automates hyperparameter tuning by iterating through experiments, trials, and training tasks.

    AutoML operates through the following mechanism:

    image

    Configure hyperparameter value ranges, search algorithms, and experiment stop conditions. AutoML creates an experiment based on this configuration.

    1. Experiments generate multiple hyperparameter combinations based on the configured algorithm. Each trial trains the model using one hyperparameter combination.

      Note

      Configure multiple trials to run concurrently to accelerate model training. This increases resource costs.

    2. Trials perform one or more computing tasks based on one hyperparameter combination. Tasks execute as DLC jobs on general computing resources or intelligent computing LINGJUN resources, or as MaxCompute tasks on MaxCompute computing resources. Billing, configuration methods, and resource usage vary between DLC jobs and MaxCompute tasks.

    3. AutoML continuously monitors task metrics during experiment execution.

    4. Experiments stop when stop conditions trigger: maximum search count reached, algorithm stop condition met, or all combinations calculated.

    5. AutoML returns results containing hyperparameter combinations or optimal models for each trial. Specify a model storage path to view models. Results are also available in logs.

    Configure these parameter types before starting experiments: basic experiment configurations, trial configurations, DLC or MaxCompute task configurations, and hyperparameter search configurations.