Watsonx.ai is IBM’s next generation enterprise studio for AI builders to train, validate, tune and deploy AI models including foundation models. This post describes briefly the available trial version.
The Watsonx.ai landing page explains the new offering:
Watsonx.ai is part of the IBM watsonx platform that brings together new generative AI capabilities, powered by foundation models, and traditional machine learning into a powerful studio spanning the AI lifecycle. With watsonx.ai, you can train, validate, tune, and deploy generative AI, foundation models, and machine learning capabilities with ease and build AI applications in a fraction of the time with a fraction of the data.
IBM Cloud provides a free trial.
For those of you who have seen the Watsonx.ai Tech Preview, you’ll notice that the Trial looks slightly different. The Tech Preview contains (obviously) the latest and greatest features, the Trial the features that are currently supported by IBM. For example, the new Tuning Studio is not part of the Trial yet.
The other difference is that in the Trial the new Foundation Models functionality is integrated in IBM’s SaaS and Cloud Pak for Data offerings. This is why you also will get three more services provisioned:
- Watson Studio
- Watson Machine Learning
- Cloud Object Storage
In the Prompt Lab, leverage foundation models to create better AI, faster. Experiment with different prompts for various use cases and tasks. With just a few lines of instruction you can draft job descriptions, classify customer complaints, summarize complex regulatory documents, extract key business information, and much more.
There are example prompts to help users to get started. Prompts can either be defined in freeform or in a more structured form, where users are guided how to write instructions and how to do few-shot prompts.
A great feature are AI guardrails:
When you toggle AI guardrails on, harmful language is automatically removed from the input prompt text as well as the output generated by the model. Specifically, any sentence in the input or output that contains harmful language will be replaced with a message saying potentially harmful text has been removed.
There is a Python API described in the documentation. When opening the right panel, you can also see how to invoke the service via REST. Note that the endpoint has a different name compared to the Tech Preview.
The following foundation models are currently supported: