Use scikit-learn to recognize hand-written digits with ibm-watsonx-ai¶

This notebook contains steps and code to demonstrate how to persist and deploy locally trained scikit-learn model in Watson Machine Learning Service. This notebook contains steps and code to work with ibm-watsonx-ai library available in PyPI repository. This notebook introduces commands for getting model and training data, persisting model, deploying model, scoring it, updating the model and redeploying it.

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

Learning goals¶

The learning goals of this notebook are:

  • Train sklearn model.
  • Persist trained model in Watson Machine Learning repository.
  • Deploy model for online scoring using client library.
  • Score sample records using client library.

Contents¶

This notebook contains the following parts:

  1. Setup
  2. Explore data and create scikit-learn model
  3. Persist externally created scikit model
  4. Deploy and score in a Cloud
  5. Clean up
  6. Summary and next steps

1. Set up the environment¶

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

  • Create a Watson Machine Learning (WML) Service instance (a free plan is offered and information about how to create the instance can be found here).

Install and import the ibm-watsonx-ai and dependecies¶

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

In [ ]:
!pip install -U ibm-watsonx-ai | tail -n 1
!pip install "scikit-learn==1.3.2" | tail -n 1

Connection to WML¶

Authenticate the Watson Machine Learning 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

In result, get the value of api_key from the output.

Location of your WML instance can be retrieved in the following way:

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

In result, get the value of location from the output.

Tip: Your Cloud API key can be generated 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 Watson Machine Learning docs. You can check your instance location in your Watson Machine Learning (WML) Service instance details.

You can also get 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 cell.

In [ ]:
api_key = 'PASTE YOUR PLATFORM API KEY HERE'
location = 'PASTE YOUR INSTANCE LOCATION HERE'
In [1]:
from ibm_watsonx_ai import Credentials

credentials = Credentials(
    api_key=api_key,
    url='https://' + location + '.ml.cloud.ibm.com'
)
In [2]:
from ibm_watsonx_ai import APIClient

client = APIClient(credentials)

Working with spaces¶

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

  • Click New Deployment Space
  • Create an empty space
  • Select Cloud Object Storage
  • Select Watson Machine Learning instance and press Create
  • Copy space_id and paste it below

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

Action: Assign space ID below

In [3]:
space_id = 'PASTE YOUR SPACE ID HERE'

You can use list method to print all existing spaces.

In [ ]:
client.spaces.list(limit=10)

To be able to interact with all resources available in Watson Machine Learning, you need to set space which you will be using.

In [4]:
client.set.default_space(space_id)
Out[4]:
'SUCCESS'

2. Explore data and create scikit-learn model¶

In this section, you will prepare and train handwritten digits model using scikit-learn library.

2.1 Explore data¶

As a first step, you will load the data from scikit-learn sample datasets and perform a basic exploration.

In [6]:
from sklearn import datasets

digits = datasets.load_digits()

Loaded toy dataset consists of 8x8 pixels images of hand-written digits.

Let's display first digit data and label using data and target.

In [7]:
print(digits.data[0].reshape((8, 8)))

[[ 0.  0.  5. 13.  9.  1.  0.  0.]
 [ 0.  0. 13. 15. 10. 15.  5.  0.]
 [ 0.  3. 15.  2.  0. 11.  8.  0.]
 [ 0.  4. 12.  0.  0.  8.  8.  0.]
 [ 0.  5.  8.  0.  0.  9.  8.  0.]
 [ 0.  4. 11.  0.  1. 12.  7.  0.]
 [ 0.  2. 14.  5. 10. 12.  0.  0.]
 [ 0.  0.  6. 13. 10.  0.  0.  0.]]
In [8]:
digits.target[0]
Out[8]:
0

In next step, you will count data examples.

In [9]:
samples_count = len(digits.images)
print("Number of samples: " + str(samples_count))

Number of samples: 1797

2.2. Create a scikit-learn model¶

Prepare data

In this step, you'll split your data into three datasets:

  • train
  • test
  • score
In [10]:
train_data = digits.data[: int(0.7*samples_count)]
train_labels = digits.target[: int(0.7*samples_count)]

test_data = digits.data[int(0.7*samples_count): int(0.9*samples_count)]
test_labels = digits.target[int(0.7*samples_count): int(0.9*samples_count)]

score_data = digits.data[int(0.9*samples_count): ]

print("Number of training records: " + str(len(train_data)))
print("Number of testing records : " + str(len(test_data)))
print("Number of scoring records : " + str(len(score_data)))

Number of training records: 1257
Number of testing records : 360
Number of scoring records : 180

Create pipeline

Next, you'll create scikit-learn pipeline.

In ths step, you will import scikit-learn machine learning packages that will be needed in next cells.

In [11]:
from sklearn.pipeline import Pipeline
from sklearn import preprocessing
from sklearn import svm, metrics

Standardize features by removing the mean and scaling to unit variance.

In [12]:
scaler = preprocessing.StandardScaler()

Next, define estimators you want to use for classification. Support Vector Machines (SVM) with radial basis function as kernel is used in the following example.

In [13]:
clf = svm.SVC(kernel='rbf')

Let's build the pipeline now. This pipeline consists of transformer and an estimator.

In [14]:
pipeline = Pipeline([('scaler', scaler), ('svc', clf)])

Train model

Now, you can train your SVM model by using the previously defined pipeline and train data.

In [15]:
model = pipeline.fit(train_data, train_labels)

Evaluate model

You can check your model quality now. To evaluate the model, use test data.

In [16]:
predicted = model.predict(test_data)

print("Evaluation report: \n\n%s" % metrics.classification_report(test_labels, predicted))

Evaluation report: 

              precision    recall  f1-score   support

           0       1.00      0.97      0.99        37
           1       0.97      0.97      0.97        34
           2       1.00      0.97      0.99        36
           3       1.00      0.94      0.97        35
           4       0.78      0.97      0.87        37
           5       0.97      0.97      0.97        38
           6       0.97      0.86      0.91        36
           7       0.92      0.97      0.94        35
           8       0.91      0.89      0.90        35
           9       0.97      0.92      0.94        37

    accuracy                           0.94       360
   macro avg       0.95      0.94      0.95       360
weighted avg       0.95      0.94      0.95       360

You can tune your model now to achieve better accuracy. For simplicity of this example tuning section is omitted.

3. Persist locally created scikit-learn model¶

In this section, you will learn how to store your model in Watson Machine Learning repository by using the IBM Watson Machine Learning SDK.

3.1: Publish model¶

Publish model in Watson Machine Learning repository on Cloud.¶

Define model name, autor name and email.

In [17]:
sofware_spec_id = client.software_specifications.get_id_by_name("runtime-24.1-py3.11")
In [18]:
metadata = {
            client.repository.ModelMetaNames.NAME: 'Scikit model',
            client.repository.ModelMetaNames.TYPE: 'scikit-learn_1.3',
            client.repository.ModelMetaNames.SOFTWARE_SPEC_ID: sofware_spec_id
}

published_model = client.repository.store_model(
    model=model,
    meta_props=metadata,
    training_data=train_data,
    training_target=train_labels)

3.2: Get model details¶

In [ ]:
import json

published_model_id = client.repository.get_model_id(published_model)
model_details = client.repository.get_details(published_model_id)
print(json.dumps(model_details, indent=2))

3.3 Get all models¶

In [ ]:
models_details = client.repository.list_models()

4. Deploy and score in a Cloud¶

In this section you will learn how to create online scoring and to score a new data record by using the IBM Watson Machine Learning SDK.

4.1: Create model deployment¶

Create online deployment for published model¶

In [21]:
metadata = {
    client.deployments.ConfigurationMetaNames.NAME: "Deployment of scikit model",
    client.deployments.ConfigurationMetaNames.ONLINE: {}
}

created_deployment = client.deployments.create(published_model_id, meta_props=metadata)


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

Synchronous deployment creation for uid: '3ea048dc-94de-4716-83de-45cbb84d0a9c' 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_uid='b7b270ac-769a-4263-b7c0-1f32bbf0d141'
------------------------------------------------------------------------------------------------

Note: Here we use deployment url saved in published_model object. In next section, we show how to retrive deployment url from Watson Machine Learning instance.

In [23]:
deployment_id = client.deployments.get_id(created_deployment)

Now you can print an online scoring endpoint.

In [ ]:
scoring_endpoint = client.deployments.get_scoring_href(created_deployment)
print(scoring_endpoint)

You can also list existing deployments.

In [ ]:
client.deployments.list()

4.2: Get deployment details¶

In [ ]:
client.deployments.get_details(deployment_id)

4.3: Score¶

You can use the following method to do test scoring request against deployed model.

Action: Prepare scoring payload with records to score.

In [25]:
score_0 = list(score_data[0])
score_1 = list(score_data[1])
In [26]:
scoring_payload = {"input_data": [{"values": [score_0, score_1]}]}

Use client.deployments.score() method to run scoring.

In [27]:
predictions = client.deployments.score(deployment_id, scoring_payload)
In [28]:
print(json.dumps(predictions, indent=2))

{
  "predictions": [
    {
      "fields": [
        "prediction"
      ],
      "values": [
        [
          5
        ],
        [
          4
        ]
      ]
    }
  ]
}

5. Clean up¶

If you want to clean up all created assets:

  • 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 scikit-learn machine learning as well as Watson Machine Learning for model creation and deployment. Check out our Online Documentation for more samples, tutorials, documentation, how-tos, and blog posts.

Authors¶

Daniel Ryszka, Software Engineer

Mateusz Szewczyk, Software Engineer at Watson Machine Learning

Copyright © 2020-2024 IBM. This notebook and its source code are released under the terms of the MIT License.