ZenML comes with caching enabled by default. As long as there is no change within a step or upstream from it, the cached outputs of that step will be used for the next pipeline run. This means that whenever there are code or configuration changes affecting a step, the step will be rerun during the next pipeline execution. Currently, the caching does not automatically detect changes within the file system or on external APIs. Make sure to set caching to
False on steps that depend on external input.