Contained within this repositoy is a set of hands-on workshops demonstrating the capabilities of Seldon Deploy across a variety of use cases and data modalities. Each workshop is a self container Jupyter notebook which, as a minimum, works through the following flow:
- Data cleaning and pre-processing
- Model training
- Deployment to Seldon
- Creation of at least one explainer, outlier detector or drift detector
The workshops are designed to be ran within a Google Colab environment. To get going all you need to do is load them into Colab, and being working through the cells.
It is likely easiest to do this with the support of the Solution Engineering team.
⚠️ Always run through the notebooks prior to using with a client: Breaking changes can happen as new versions are released!
- Adapt notebooks so they can be easily ran locally
- Ensure they are all up to date
- Make it easier to update cluster IP
- Better instructions for use.