Watson Machine Learning Python client samples and examples
Review and use sample Jupyter Notebooks that use Watson Machine Learning Python library to demonstrate machine learning features and techniques. Each notebook lists learning goals so you can find the one that best meets your goals.
Training and deploying models from notebooks
If you choose to build a machine learning model in a notebook, you must be comfortable with coding in a Jupyter Notebook. A Jupyter Notebook is a web-based environment for interactive computing. You can run small pieces of code that process your data, and then immediately view the results of your computation. Using this tool, you can assemble, test, and run all of the building blocks you need to work with data, save the data to Watson Machine Learning, and deploy the model.
Learn from sample notebooks
Many ways exist to build and train models and then deploy them. Therefore, the best way to learn is to look at annotated samples that step you through the process by using different frameworks. Review representative samples that demonstrate key features.
The samples are built by using the V4 version of the watsonx.ai Python client library.
Watch this video to learn how to train, deploy, and test a machine learning model in a Jupyter Notebook. This video mirrors the Use scikit-learn to recognize hand-written digits found in the Deployment samples table.
This video provides a visual method to learn the concepts and tasks in this documentation.
Helpful variables
Use the pre-defined PROJECT_ID environment variable to call the watsonx.ai Python client APIs. PROJECT_ID is the guide of the project where your environment is running.
Deployment samples
View or run these Jupyter Notebooks to see how techniques are implemented by using various frameworks. Some of the samples rely on trained models, which are also available for you to download from the public repository.
| Sample name | Framework | Techniques demonstrated |
|---|---|---|
| Use scikit-learn and custom library to predict temperature | Scikit-learn | Train a model with custom defined transformer Persist the custom-defined transformer and the model in Watson Machine Learning repository Deploy the model by using Watson Machine Learning Service Perform predictions that use the deployed model |
| Use scikit-learn and AI lifecycle capabilities to predict California house prices with ibm-watsonx-ai | Scikit-learn | Download an externally trained scikit-learn model with dataset Persist an external model in Watson Machine Learning repository Deploy model for online scoring using client library Score sample records using client library Update previously persisted model Redeploy model in-place Scale deployment |
| Use PMML to predict iris species | PMML | Deploy and score a PMML model |
| Use Python function to recognize hand-written digits | Python | Use a function to store a sample model, then deploy the sample model. |
| Use scikit-learn to recognize hand-written digits | Scikit-learn | Train sklearn model Persist trained model in Watson Machine Learning repository Deploy model for online scoring by using client library Score sample records by using client library |
| Use Spark and batch deployment to predict customer churn | Spark | Load a CSV file into an Apache Spark DataFrame Explore data Prepare data for training and evaluation Create an Apache Spark machine learning pipeline Train and evaluate a model Persist a pipeline and model in Watson Machine Learning repository Explore and visualize prediction result by using the plotly package Deploy a model for batch scoring by using Watson Machine Learning API |
| Use Spark and Python to predict Credit Risk | Spark | Load a CSV file into an Apache® Spark DataFrame Explore data Prepare data for training and evaluation Persist a pipeline and model in Watson Machine Learning repository from tar.gz files Deploy a model for online scoring by using Watson Machine Learning API Score sample data by using the Watson Machine Learning API Explore and visualize prediction results by using the plotly package. |
| Use SPSS to predict customer churn | SPSS | Work with the instance Perform an online deployment of the SPSS model Score data by using deployed model |
| Use XGBoost to classify tumors | XGBoost | Load a CSV file into numpy array Explore data Prepare data for training and evaluation Create an XGBoost machine learning model Train and evaluate a model Use cross-validation to optimize the model's hyperparameters Persist a model in Watson Machine Learning repository Deploy a model for online scoring Score sample data |
| Use Keras to recognize hand-written digits.ipynb | Keras (Tensorflow) | Download an externally trained Keras model with dataset. Persist an external model in the Watson Machine Learning repository. Deploy a model for online scoring by using client library. Score sample records by using client library. |
| Machine Learning artifact management | Core | Export and import artifacts Load, deploy, and score externally created models |
| Use custom image, software specification and runtime definition to deploy a python function with ibm-watsonx-ai | Python | Creating a Custom Image Creating custom software specification and runtime definition Creating an online deployment for Python function Scoring data using deployed function |
| Use R Shiny app to create SIR model | R | Persist an R Shiny app in Watson Machine Learning asset repository. Deploy an application for online scoring by using client library. Score sample records by using client library. |
Use watsonx, and meta-llama/llama-3-1-8b-instruct to run as an AI service |
Python | Setup Create AI service Testing AI service's function locally Deploy AI service Example of Executing an AI service Summary |
| Use watsonx, and meta-llama/Meta-Llama-3-8B to Fine Tune with online banking queries annotated | Python | Setup Data loading Initialize experiment Run Fine Tuning Fine Tuning details List historical Fine Tuning experiments Deploy Tuned Model Preparing data for testing process Foundation Models Inference on watsonx.ai Deploy Base Model Comparison of prediction results Clean up Summary and next steps |
AutoAI samples
View or run these Jupyter Notebooks to see how AutoAI model techniques are implemented.
| Sample name | Framework | Techniques demonstrated |
|---|---|---|
| Use AutoAI and Lale to predict credit risk | Hybrid (AutoAI) with Lale | Work with Watson Machine Learning experiments to train AutoAI models. Compare trained models quality and select the best one for further refinement Refine the best model and test new variations Deploy and score the trained model |
| Use AutoAI RAG and Chroma to create a pattern and get information from ibm-watsonx-ai SDK documentation | Setup RAG Optimizer definition RAG Experiment run RAG Patterns comparison and testing Historical runs Clean up Summary and next steps |
|
| Use AutoAI RAG and Milvus database to work with ibm-watsonx-ai SDK documentation | Setup RAG Optimizer definition RAG Experiment run RAG Patterns comparison and testing Historical runs Clean up Summary and next steps |
More samples
View or download sample notebooks from the GitHub repository to use as a model for your tasks. Most of the notebooks are annotated to describe key actions and features.
Next steps
- To learn more about using notebook editors, see Notebooks.
- To learn more about working with notebooks, see Coding and running notebooks.
Parent topic: Managing predictive deployments