AI SDK, Tutorials and guides
End-to-end tutorials, how-to Jupyter notebooks and guides for Siemens' AI Software Development Kit, which packages, tests and creates inference pipelines for deployment on the AI Inference Server.
AI SDK is a set of Python libraries that provide building blocks for packaging a trained model and its dependencies, verifying and testing that package locally, and creating an inference pipeline package to run the model on Siemens' AI Inference Server; it is the entry point into Siemens' Industrial AI portfolio and is published on PyPI as simaticai.
This repository's end-to-end tutorials walk through a full notebook-based machine learning workflow: preparing training data, training an example model, packaging it as an inference pipeline, testing the packaged pipeline, and generating the pipeline for AI@Edge deployment. Separate how-to guides and how-to notebooks cover specific aspects of packaging a model for the AI Inference Server and can be used as a starting point for packaging a reader's own models.
From AI SDK version 2.6.0 onward, the tutorials are tested only on Python 3.12 on Linux-type systems, though other environments may still work.