.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "_examples/generative_design_ex/01-build_model.py"
.. LINE NUMBERS ARE GIVEN BELOW.

.. only:: html

    .. note::
        :class: sphx-glr-download-link-note

        :ref:`Go to the end <sphx_glr_download__examples_generative_design_ex_01-build_model.py>`
        to download the full example code.

.. rst-class:: sphx-glr-example-title

.. _sphx_glr__examples_generative_design_ex_01-build_model.py:

.. _ref_build_model:

Building a Generative Design Model
================================================

This example demonstrates how to configure a Generative Design model, start the model training process, and monitor the build progress.

Before you begin
-------------------------------------------

- Complete ":ref:`ref_create_project_upload_data`" to create a project with training data.
- Ensure all training data in your project are ready (processed successfully).

.. GENERATED FROM PYTHON SOURCE LINES 39-41

Import necessary libraries
-------------------------------------------

.. GENERATED FROM PYTHON SOURCE LINES 41-45

.. code-block:: Python


    import ansys.simai.core as asc
    from ansys.simai.core.data.geomai.models import GeomAIModelConfiguration


.. GENERATED FROM PYTHON SOURCE LINES 46-49

Configure your settings
-------------------------------------------
Update these variables with your specific settings:

.. GENERATED FROM PYTHON SOURCE LINES 49-55

.. code-block:: Python


    ORGANIZATION = "my_organization"  # Replace with your organization name
    PROJECT_NAME = "new-bracket-project"  # Replace with your project name
    NB_LATENT_PARAMS = 15  # Number of latent parameters (2-256)
    BUILD_PRESET = "default"  # Options: "debug", "short", "default", "long"


.. GENERATED FROM PYTHON SOURCE LINES 56-62

The ``BUILD_PRESET`` options correspond to:

- ``"debug"``: Fast training for testing (very few epochs).
- ``"short"``: Quick training with reduced accuracy.
- ``"default"``: Balanced training time and quality.
- ``"long"``: Longer training for best quality.

.. GENERATED FROM PYTHON SOURCE LINES 65-68

Initialize the client and get the project
-------------------------------------------
Connect to the instance:

.. GENERATED FROM PYTHON SOURCE LINES 68-72

.. code-block:: Python


    simai_client = asc.SimAIClient(organization=ORGANIZATION)
    geomai_client = simai_client.geomai


.. GENERATED FROM PYTHON SOURCE LINES 73-74

Retrieve the project by name:

.. GENERATED FROM PYTHON SOURCE LINES 74-78

.. code-block:: Python


    project = geomai_client.projects.get(name=PROJECT_NAME)
    print(f"Using project: {project.name}")


.. GENERATED FROM PYTHON SOURCE LINES 79-82

Verify project data are ready
-------------------------------------------
Before building a model, ensure all training data are processed:

.. GENERATED FROM PYTHON SOURCE LINES 82-90

.. code-block:: Python


    project_data = project.data()
    ready_data = [data for data in project_data if data.is_ready]
    print(f"Project has {len(ready_data)}/{len(project_data)} ready training data")

    if len(ready_data) < len(project_data):
        print("Warning: Some training data are not ready. Model may fail to build.")


.. GENERATED FROM PYTHON SOURCE LINES 91-98

Configure the model
-------------------------------------------
To define the configuration for your model, you can either specify the build preset or the number
of epochs.
To do so, instead of ``build_preset``, you can specify the number of epochs directly. Example: ``nb_epochs=100``.

The number of latent parameters defines the complexity of the model's latent space; start with a small number (e.g., 10) and adjust based on your needs.

.. GENERATED FROM PYTHON SOURCE LINES 98-105

.. code-block:: Python



    configuration = GeomAIModelConfiguration(
        build_preset=BUILD_PRESET,
        nb_latent_param=NB_LATENT_PARAMS,  # Must be between 2 and 256
    )


.. GENERATED FROM PYTHON SOURCE LINES 106-109

Build the model
-------------------------------------------
Start the model training process:

.. GENERATED FROM PYTHON SOURCE LINES 109-113

.. code-block:: Python


    model = geomai_client.models.build(project, configuration)
    print(f"Started build for model {model.id}")


.. GENERATED FROM PYTHON SOURCE LINES 114-117

Wait for model training to complete
-------------------------------------------
Monitor the training process and handle completion:

.. GENERATED FROM PYTHON SOURCE LINES 117-127

.. code-block:: Python


    print("Waiting for model training to complete (this may take a while)...")
    while not model.wait(timeout=60):  # Check every 60 seconds
        print(f"Model {model.id} is still training...")

    if model.has_failed:
        print(f"Model training failed: {model.failure_reason}")
    else:
        print(f"Model {model.id} trained successfully and is ready to use.")


.. GENERATED FROM PYTHON SOURCE LINES 128-134

Next steps
-------------------------------------------
Once your model is trained, you can:

- Generate random geometries: :ref:`ref_generate_random_geometries`
- Interpolate between geometries: :ref:`ref_interpolate_geometries`


.. _sphx_glr_download__examples_generative_design_ex_01-build_model.py:

.. only:: html

  .. container:: sphx-glr-footer sphx-glr-footer-example

    .. container:: sphx-glr-download sphx-glr-download-jupyter

      :download:`Download Jupyter notebook: 01-build_model.ipynb <01-build_model.ipynb>`

    .. container:: sphx-glr-download sphx-glr-download-python

      :download:`Download Python source code: 01-build_model.py <01-build_model.py>`

    .. container:: sphx-glr-download sphx-glr-download-zip

      :download:`Download zipped: 01-build_model.zip <01-build_model.zip>`


.. only:: html

 .. rst-class:: sphx-glr-signature

    `Gallery generated by Sphinx-Gallery <https://sphinx-gallery.github.io>`_