GitHub Repository: IBM/watson-machine-learning-samples
Path: blob/master/cloud/notebooks/python_sdk/converters/Use ONNX model converted from AutoAI.ipynb
⁵⁰⁸⁷ views

Kernel: .venv_samples_py_311_x86

Use ONNX model converted from AutoAI with `ibm-watsonx-ai`

This notebook facilitates ONNX, AutoAI, and watsonx.ai Runtime service. It contains steps and code to work with ibm-watsonx-ai library available in PyPI repository in order to convert the model to ONNX format. It also introduces commands for persisting, deploying and scoring the model.

Some familiarity with Python is helpful. This notebook uses Python 3.11.

Learning goals

The learning goals of this notebook are:

Train an AutoAI model
Convert the native scikit-learn model to ONNX format
Deploy the model for online scoring using client library
Score sample records using the client library

This notebook contains the following parts:

1. Environment setup

Before you use the sample code in this notebook, you must perform the following setup tasks:

Create a watsonx.ai Runtime instance (information on service plans and further reading can be found here).

1.1. Installing and importing the `ibm-watsonx-ai` and dependencies

Note: ibm-watsonx-ai documentation can be found here.

In [1]:

%pip install ibm-watsonx-ai | tail -n 1
%pip install wget | tail -n 1
%pip install "autoai_libs[onnx]<3.0.0" | tail -n 1
%pip install onnxruntime | tail -n 1

Out[1]:

Successfully installed anyio-4.12.0 cachetools-6.2.2 certifi-2025.11.12 charset_normalizer-3.4.4 h11-0.16.0 httpcore-1.0.9 httpx-0.28.1 ibm-cos-sdk-2.14.3 ibm-cos-sdk-core-2.14.3 ibm-cos-sdk-s3transfer-2.14.3 ibm-watsonx-ai-1.4.11 idna-3.11 jmespath-1.0.1 lomond-0.3.3 numpy-2.3.5 pandas-2.2.3 pytz-2025.2 requests-2.32.5 tabulate-0.9.0 tzdata-2025.2 urllib3-2.6.2
Successfully installed wget-3.2
Successfully installed astunparse-1.6.3 attrs-25.4.0 autoai_libs-2.0.29 black-25.12.0 click-8.3.1 cloudpickle-3.1.2 future-1.0.0 gensim-4.3.3 graphviz-0.21 greenery-3.3.3 hyperopt-0.2.7 joblib-1.5.2 jsonref-1.1.0 jsonschema-4.20.0 jsonschema-specifications-2025.9.1 jsonsubschema-0.0.7 lale-0.8.4 lightgbm-4.2.0 mypy-extensions-1.1.0 networkx-3.6.1 numpy-1.26.4 onnx-1.16.0 onnxconverter-common-1.13.0 onnxmltools-1.14.0 onnxruntime-extensions-0.13.0 pandas-2.1.4 parameterized-0.8.1 pathspec-0.12.1 portion-2.6.1 protobuf-6.33.2 py4j-0.10.9.9 pytokens-0.3.0 referencing-0.37.0 rpds-py-0.30.0 scikit-learn-1.3.2 scipy-1.13.1 skl2onnx-1.18.0 smart-open-7.5.0 sortedcontainers-2.4.0 threadpoolctl-3.6.0 tqdm-4.67.1 wheel-0.45.1 wrapt-2.0.1 xgboost-2.0.3
Successfully installed coloredlogs-15.0.1 flatbuffers-25.9.23 humanfriendly-10.0 mpmath-1.3.0 onnxruntime-1.23.2 sympy-1.14.0

1.2. Connecting to watsonx.ai Runtime

Authenticate with the watsonx.ai Runtime service on IBM Cloud. You need to provide platform api_key and instance location.

You can use IBM Cloud CLI to retrieve platform API Key and instance location.

API Key can be generated in the following way:

ibmcloud login
ibmcloud iam api-key-create API_KEY_NAME

Get the value of api_key from the output.

Location of your watsonx.ai Runtime instance can be retrieved in the following way:

ibmcloud login --apikey API_KEY -a https://cloud.ibm.com
ibmcloud resource service-instance INSTANCE_NAME

Get the value of location from the output.

Tip: You can generate your Cloud API key by going to the Users section of the Cloud console. From that page, click your name, scroll down to the API Keys section, and click Create an IBM Cloud API key. Give your key a name and click Create, then copy the created key and paste it below. You can also get a service-specific url by going to the Endpoint URLs section of the watsonx.ai Runtime docs. You can check your instance location in your watsonx.ai Runtime Service instance details.

You can also get the service specific apikey by going to the Service IDs section of the Cloud Console. From that page, click Create, then copy the created key, and paste it below.

Action: Enter your api_key and location in the following cells.

In [2]:

import getpass

api_key = getpass.getpass("Please enter your api key (hit enter): ")

In [3]:

location = "PASTE YOUR LOCATION HERE"

If you are running this notebook on Cloud, you can access the location via:

location = os.environ.get("RUNTIME_ENV_REGION")

In [4]:

from ibm_watsonx_ai import Credentials

credentials = Credentials(api_key=api_key, url=f"https://{location}.ml.cloud.ibm.com")

In [5]:

from ibm_watsonx_ai import APIClient

client = APIClient(credentials=credentials)

1.3. Working with spaces

First of all, you need to create a space that will be used for your work. If you do not have a space, you can use Deployment Spaces Dashboard to create one.

Click New Deployment Space
Create an empty space
Select Cloud Object Storage
Select watsonx.ai Runtime instance and press Create
Copy space_id and paste it below

Tip: You can also use the ibm_watsonx_ai SDK to prepare the space for your work. More information can be found here.

Action: Assign space ID below

In [6]:

space_id = "PASTE YOUR SPACE ID HERE"

You can use the list method to print all existing spaces.

In [ ]:

client.spaces.list(limit=10)

To be able to interact with all resources available in watsonx.ai Runtime, you need to set space which you will be using.

In [7]:

client.set.default_space(space_id)

Out[7]:

'SUCCESS'

Connections to COS

In next cell we read the COS credentials from the space.

In [8]:

from ibm_watsonx_ai.utils import get_from_json

In [9]:

space_details = client.spaces.get_details(space_id=space_id)

cos_credentials = get_from_json(space_details, ["entity", "storage", "properties"])

2. Optimizer definition

Training data connection

Define connection information to COS bucket and training data CSV file. This example uses the German Credit Risk dataset.

The code in next cell uploads training data to the bucket.

In [10]:

filename = "credit_risk_training_light.csv"
datasource_name = "bluemixcloudobjectstorage"
bucket_name = cos_credentials.get("bucket_name")

Download training data from git repository.

In [11]:

import os

import wget

url = "https://raw.githubusercontent.com/IBM/watsonx-ai-samples/master/cloud/data/credit_risk/credit_risk_training_light.csv"
if not os.path.isfile(filename):
    wget.download(url)

Create connection

In [12]:

conn_meta_props = {
    client.connections.ConfigurationMetaNames.NAME: f"Connection to Database - {datasource_name} ",
    client.connections.ConfigurationMetaNames.DATASOURCE_TYPE: client.connections.get_datasource_type_id_by_name(
        datasource_name
    ),
    client.connections.ConfigurationMetaNames.DESCRIPTION: "Connection to external Database",
    client.connections.ConfigurationMetaNames.PROPERTIES: {
        "bucket": bucket_name,
        "access_key": get_from_json(
            cos_credentials, ["credentials", "editor", "access_key_id"]
        ),
        "secret_key": get_from_json(
            cos_credentials, ["credentials", "editor", "secret_access_key"]
        ),
        "iam_url": "https://iam.cloud.ibm.com/identity/token",
        "url": cos_credentials.get("endpoint_url"),
    },
}

conn_details = client.connections.create(meta_props=conn_meta_props)

Out[12]:

Creating connections...
SUCCESS

Note: The above connection can be initialized alternatively with api_key and resource_instance_id. The above cell can be replaced with:

conn_meta_props= {
    client.connections.ConfigurationMetaNames.NAME: f"Connection to Database - {db_name} ",
    client.connections.ConfigurationMetaNames.DATASOURCE_TYPE: client.connections.get_datasource_type_id_by_name(db_name),
    client.connections.ConfigurationMetaNames.DESCRIPTION: "Connection to external Database",
    client.connections.ConfigurationMetaNames.PROPERTIES: {
        "bucket": bucket_name,
        "api_key": cos_credentials["apikey"],
        "resource_instance_id": cos_credentials["resource_instance_id"],
        "iam_url": "https://iam.cloud.ibm.com/identity/token",
        "url": "https://s3.us.cloud-object-storage.appdomain.cloud"
    }
}

conn_details = client.connections.create(meta_props=conn_meta_props)

In [13]:

connection_id = client.connections.get_id(conn_details)

Define connection information to training data.

In [14]:

from ibm_watsonx_ai.helpers import DataConnection, S3Location

credit_risk_conn = DataConnection(
    connection_asset_id=connection_id,
    location=S3Location(bucket=bucket_name, path=filename),
)

training_data_reference = [credit_risk_conn]

Check the connection information. Upload the data and validate.

In [15]:

credit_risk_conn.set_client(client)
credit_risk_conn.write(data=filename, remote_name=filename)
credit_risk_conn.read()

Out[15]:

Optimizer configuration

Provide the input information for AutoAI optimizer:

name - experiment name
prediction_type - type of the problem
prediction_column - target column name
scoring - optimization metric

In [16]:

from ibm_watsonx_ai.experiment import AutoAI

experiment = AutoAI(credentials, space_id=space_id)

pipeline_optimizer = experiment.optimizer(
    name="Credit Risk Prediction - AutoAI",
    prediction_type=AutoAI.PredictionType.BINARY,
    prediction_column="Risk",
    include_only_estimators=["RandomForestClassifier"],
)

Configuration parameters can be retrieved via get_params().

In [17]:

pipeline_optimizer.get_params()

Out[17]:

{'name': 'Credit Risk Prediction - AutoAI',
 'desc': '',
 'prediction_type': 'binary',
 'prediction_column': 'Risk',
 'prediction_columns': None,
 'timestamp_column_name': None,
 'scoring': None,
 'holdout_size': None,
 'max_num_daub_ensembles': None,
 't_shirt_size': 'l',
 'train_sample_rows_test_size': None,
 'include_only_estimators': [<ClassificationAlgorithms.RF: 'RandomForestClassifier'>],
 'include_batched_ensemble_estimators': None,
 'backtest_num': None,
 'lookback_window': None,
 'forecast_window': None,
 'backtest_gap_length': None,
 'cognito_transform_names': None,
 'csv_separator': ',',
 'excel_sheet': None,
 'encoding': 'utf-8',
 'positive_label': None,
 'drop_duplicates': True,
 'outliers_columns': None,
 'text_processing': None,
 'word2vec_feature_number': None,
 'daub_give_priority_to_runtime': None,
 'text_columns_names': None,
 'sampling_type': None,
 'sample_size_limit': None,
 'sample_rows_limit': None,
 'sample_percentage_limit': None,
 'number_of_batch_rows': None,
 'n_parallel_data_connections': None,
 'test_data_csv_separator': ',',
 'test_data_excel_sheet': None,
 'test_data_encoding': 'utf-8',
 'categorical_imputation_strategy': None,
 'numerical_imputation_strategy': None,
 'numerical_imputation_value': None,
 'imputation_threshold': None,
 'retrain_on_holdout': True,
 'feature_columns': None,
 'pipeline_types': None,
 'supporting_features_at_forecast': None,
 'numerical_columns': None,
 'categorical_columns': None,
 'confidence_level': None,
 'incremental_learning': None,
 'early_stop_enabled': None,
 'early_stop_window_size': None,
 'time_ordered_data': None,
 'feature_selector_mode': None,
 'run_id': None}

3. Experiment run

Call the fit() method to trigger the AutoAI experiment. You can either use interactive mode (synchronous job) or background mode (asychronous job) by specifying background_model=True.

In [18]:

run_details = pipeline_optimizer.fit(
    training_data_reference=training_data_reference, background_mode=False
)

Out[18]:

Training job c5bb1f49-573e-49d9-8641-2a35adb5aa35 completed: 100%|████████| [03:03<00:00,  1.84s/it]

You can use the get_run_status() method to monitor AutoAI jobs in background mode.

In [19]:

pipeline_optimizer.get_run_status()

Out[19]:

'completed'

In [20]:

pipeline_optimizer.summary()

Out[20]:

In [21]:

pipeline_name = "Pipeline_1"

In [22]:

pipeline_model = pipeline_optimizer.get_pipeline(
    pipeline_name=pipeline_name, astype=AutoAI.PipelineTypes.ONNX
)

3.1. Model evaluation

In [23]:

X_test = credit_risk_conn.read().drop(["Risk"], axis=1)[:3]

In [24]:

X_test_dict = {col: X_test[col].apply(lambda x: [x]).tolist() for col in X_test.columns}
pipeline_model.run([], X_test_dict)

Out[24]:

[array(['No Risk', 'No Risk', 'No Risk'], dtype=object),
 [{'No Risk': 0.8999999761581421, 'Risk': 0.10000000149011612},
  {'No Risk': 1.0, 'Risk': 0.0},
  {'No Risk': 0.800000011920929, 'Risk': 0.20000000298023224}]]

4. Deploy and score

In this section you will learn how to deploy and score pipeline model as webservice using watsonx.ai Runtime instance.

Online deployment creation

In [25]:

run_id = get_from_json(run_details, ["metadata", "id"])

In [26]:

from ibm_watsonx_ai.deployment import WebService

service = WebService(credentials, source_space_id=space_id)

service.create(
    experiment_run_id=run_id,
    model=pipeline_name,
    deployment_name=f"Credit Risk Deployment AutoAI - ONNX - {pipeline_name}",
)

Out[26]:

Preparing an AutoAI Deployment...
Published model uid: 62e3bf7b-37f7-4c38-8e4e-db82ce5a9f79
Deploying model 62e3bf7b-37f7-4c38-8e4e-db82ce5a9f79 using V4 client.


######################################################################################

Synchronous deployment creation for id: '62e3bf7b-37f7-4c38-8e4e-db82ce5a9f79' started

######################################################################################


initializing
Note: online_url and serving_urls are deprecated and will be removed in a future release. Use inference instead.
.....
ready


-----------------------------------------------------------------------------------------------
Successfully finished deployment creation, deployment_id='368c2d8f-cb12-430f-be1d-0ff493b2d3d5'
-----------------------------------------------------------------------------------------------

Deployment object could be printed to show basic information:

In [27]:

print(service)

To show all available information about the deployment use the .get_params() method:

In [28]:

service.get_params()

In [29]:

deployment_id = get_from_json(service.get_params(), ["metadata", "id"])
deployment_id

Out[29]:

Note: online_url and serving_urls are deprecated and will be removed in a future release. Use inference instead.

'368c2d8f-cb12-430f-be1d-0ff493b2d3d5'

In [30]:

scoring_payload = {"input_data": [{"values": X_test}]}

Webservice scoring

You can make scoring request by calling score() on deployed pipeline.

In [31]:

predictions = client.deployments.score(deployment_id, scoring_payload)

predictions

Out[31]:

{'predictions': [{'fields': ['prediction', 'probability'],
   'values': [['No Risk', [0.9, 0.1]],
    ['No Risk', [1.0, 0.0]],
    ['No Risk', [0.8, 0.2]]]}]}

If you want to work with the web service in an external Python application you can retrieve the service object by:

Initialize the service by service = WebService(credentials)
Get deployment_id by service.list() method
Get webservice object by service.get('deployment_id') method

After that you can call service.score() method.

Deleting deployment

You can delete the existing deployment by calling the service.delete() command. To list the existing web services you can use service.list().

5. Cleanup

If you want to clean up after the notebook execution, i.e. remove any created assets like:

experiments
trainings
pipelines
model definitions
models
functions
deployments

please follow up this sample notebook.

6. Summary and next steps

You successfully completed this notebook! You learned how to use ONNX, scikit-learn machine learning library as well as watsonx.ai Runtime for model creation and deployment. Check out our Online Documentation for more samples, tutorials, documentation, how-tos, and blog posts.

Authors

Marta Tomzik, Software Engineer at watsonx.ai

Use ONNX model converted from AutoAI with `ibm-watsonx-ai`

Learning goals

Contents

1. Environment setup

1.1. Installing and importing the `ibm-watsonx-ai` and dependencies

1.2. Connecting to watsonx.ai Runtime

1.3. Working with spaces

Connections to COS

2. Optimizer definition

Training data connection

Create connection

Optimizer configuration

3. Experiment run

3.1. Model evaluation

4. Deploy and score

Online deployment creation

Webservice scoring

Deleting deployment

5. Cleanup

6. Summary and next steps

Authors

Product

Resources

Company

Use ONNX model converted from AutoAI with ibm-watsonx-ai

Learning goals

Contents

1. Environment setup

1.1. Installing and importing the ibm-watsonx-ai and dependencies

1.2. Connecting to watsonx.ai Runtime

1.3. Working with spaces

Connections to COS

2. Optimizer definition

Training data connection

Create connection

Optimizer configuration

3. Experiment run

3.1. Model evaluation

4. Deploy and score

Online deployment creation

Webservice scoring

Deleting deployment

5. Cleanup

6. Summary and next steps

Authors

Use ONNX model converted from AutoAI with `ibm-watsonx-ai`

1.1. Installing and importing the `ibm-watsonx-ai` and dependencies