IBM watsonx.ai

Version: 9.3.0    Premium   IBM

Thumbnail depiction of the interface of this serviceDetails for one of the IBM Granite foundation models. Thumbnail depiction of the interface of this serviceThe Prompt Lab is opened in chat mode. Thumbnail depiction of the interface of this serviceSome of the available foundation models are displayed.

Description

Train, validate, tune, and deploy generative AI solutions with foundation models in IBM watsonx.ai.

In watsonx.ai, you can use large language models from IBM and other providers. The available foundation models support a range of use cases for both natural languages and programming languages.

From the full watsonx.ai service, experiment with prompt engineering in the Prompt Lab, a tool that is designed to help you prompt foundation models. Use built-in sample prompts to get started with confidence.

Store effective prompts as prompt template assets that you can reuse and share with others. Or store the prompt as a notebook asset. The prompt text, model reference, and prompt engineering parameters are formatted as Python code in a notebook that you can interact with programmatically.

Use the Tuning Studio to guide a foundation model to return output that better meets your needs.

Quick links

Integrated services

Table 1. Supplemental services. You can extend the functionality of this service with the following supplemental services, which require this service.
Service Capability
Analytics Engine powered by Apache Spark Run analytical, machine learning, and Spark API jobs on Apache Spark clusters.
SPSS Modeler Create flows to prepare data, develop and manage models, and visualize data. No coding required.
Watson Machine Learning Build, train, and deploy machine learning models with a full range of tools.
Decision Optimization Find the most appropriate prescriptive solutions to your business problems by using CPLEX optimization engines to evaluate millions of possibilities.
RStudio® Server Runtimes Access the RStudio IDE.
Execution Engine for Apache Hadoop Integrate the Watson Studio service with your remote Apache Hadoop cluster so you can explore data and build and deploy models on your remote cluster.
Orchestration Pipelines Use Orchestration Pipelines and create end-to-end flows of machine learning pipelines to create models and customize various functions.
Table 2. Related services. The following related services are often used with this service and provide complementary features, but they are not required.
Service Capability
watsonx.governance Accelerate responsible, transparent, and explainable AI workflows for generative AI models and machine learning models.
Watson Machine Learning Accelerator Manage the lifecycle of training Deep Learning models, from data ingest and preparation to moving the model into production.