DOCSEARCHbyAlgolia
Submit Docs

IBM watsonx.ai Samples

repository·master·Indexed 41 days ago

https://github.com/ibm/watsonx-ai-samples

A collection of sample notebooks, models, and applications demonstrating the capabilities of IBM watsonx.ai and Watson Machine Learning. Includes practical guides for model building, deployment, monitoring, and foundation model integration using Python SDKs and REST APIs. Features examples for AutoAI, Deep Learning, RAG with LangChain, and integration with Watson OpenScale for bias and drift detection.

Tokens
51.4K
Snippets
144
Records
195
Agent score
47%

What's inside watsonx.ai

  1. Overview of watsonx.ai sample notebooks

    master

    This repository provides sample notebooks demonstrating various capabilities of Watson Machine Learning and watsonx.ai for both IBM Cloud and the watsonx.ai software product.

    Key capabilities covered include:

    • Model Building: Running experiments using AutoAI or Deep Learning.
    • Model Deployment: Deploying third-party models (e.g., scikit-learn, xgboost, keras, PMMl, SPSS) as web services or batch jobs.
    • Model Monitoring: Using OpenScale for drift and bias detection.
    • Lifecycle Management: Updating model versions and refreshing deployments.
    • Foundation Models: Inferencing watsonx.ai foundation models.
    • Orchestration: Integrating LangChain with watsonx.ai.

  2. Overview of watsonx.ai Cloud samples

    master
    This repository contains sample models, notebooks, and applications specifically designed for Watson Machine Learning within the IBM watsonx.ai ecosystem (both for IBM Cloud and the watsonx.ai software product). These samples serve as reference implementations for developers looking to integrate and use watsonx.ai capabilities.

  4. Explore neural network architectures and ML pipelines

    master
    The cpd3.5/definitions directory contains source code examples for deep learning neural network architectures and machine learning pipelines. Developers can use these samples as templates or reference implementations for building neural networks and orchestrating ML workflows within the watsonx.ai environment.

  5. Explore Watson Machine Learning sample models

    master

    The cpd5.1/models directory contains sample models designed to demonstrate the core capabilities of Watson Machine Learning. These samples are intended to help users understand how to implement and manage machine learning workflows within the product, specifically focusing on:

    • Deployment: How to take a trained model and make it available for inference.
    • Lifecycle Management: Managing the various stages of a model's existence, from creation to retirement.
    • Monitoring: Tracking model performance and health in production environments.

  7. Explore sample neural network architectures and ML pipelines

    master
    The cpd5.3/definitions directory contains source code examples for deep learning neural network architectures and machine learning pipelines. Developers can use these samples as templates or reference implementations for building complex model definitions and automated workflows within the watsonx.ai ecosystem.

  9. Explore sample models for Watson Machine Learning

    master

    The cpd4.0/models directory contains sample models designed to demonstrate the core capabilities of Watson Machine Learning. These samples are intended to show how to implement and manage machine learning workflows, specifically focusing on:

    • Deployment: How to take a trained model and make it available for inference.
    • Lifecycle Management: Managing the stages of a model from creation through updates.
    • Monitoring: Tracking model performance and health in production environments.

  10. Explore watsonx.ai sample notebooks

    master

    The cpd5.2/notebooks directory contains sample notebooks for demonstrating watsonx.ai capabilities. Depending on your preferred interaction method, you can find examples in two main subdirectories:

    • Python SDK: Notebooks containing Python code and utilizing the Python SDK are located in the python_sdk folder.
    • REST API: Examples demonstrating interactions via REST API are located in the rest_api folder.

  11. Explore watsonx.ai sample notebooks

    master

    This repository contains sample notebooks for IBM watsonx.ai (both for IBM Cloud and the software product) to demonstrate Watson Machine Learning capabilities. The notebooks are categorized by functional area:

    • Experiments: Triggering AutoAI or Deep Learning experiments, comparing models, and performing refinery.
    • Deployments: Serving models as online web services or batch jobs using frameworks like scikit-learn, xgboost, keras, tensorflow, pytorch, pmml, SPSS, spark MLlib, or custom python functions.
    • Monitoring: End-to-end scenarios for monitoring production models using Watson Machine Learning and Watson OpenScale to detect bias and drift.
    • Lifecycle Management: Updating existing model versions and refreshing existing deployments in-place.

    Depending on your preferred interface, notebooks are organized into two main sub-directories:

    • python_sdk: Contains notebooks using the Python SDK.
    • rest_api: Contains notebooks using REST API calls.

  12. Explore Watson Machine Learning sample notebooks

    master

    The cpd4.7/notebooks directory contains sample notebooks for demonstrating Watson Machine Learning capabilities. These notebooks are categorized by functional area:

    • experiments: Triggering AutoAI or Deep Learning experiments, working with trained models, model comparison, refinery, and deployment.
    • deployments: Serving models as online web services or batch jobs using frameworks like Spark MLlib, PMML, SPSS, scikit-learn, xgboost, keras, tensorflow, pytorch, or custom python functions.
    • monitoring: End-to-end scenarios for monitoring models in production using Watson Machine Learning and Watson OpenScale to detect bias and drift.
    • lifecycle management: Updating existing model versions and refreshing existing deployments in-place.

    Notebooks are provided in two formats:

    • Python SDK: Located in the python_sdk folder.
    • REST API: Located in the rest_api folder.