This eliminates the need to re-download datasets during each setup job (as this is slow) - we can temporarily cache datasets for the builds during the week and only re-download them once a week. Datasets will not be available for public distribution, i.e. we are not hosting them, but caching them internally for the builds.
The ask is Caching - caching on the build machine, i.e. download if does not exist
This eliminates the need to re-download datasets during each setup job (as this is slow) - we can temporarily cache datasets for the builds during the week and only re-download them once a week. Datasets will not be available for public distribution, i.e. we are not hosting them, but caching them internally for the builds.
The ask is Caching - caching on the build machine, i.e. download if does not exist