[ENH] add caching action for elephant-data

[Moritz-Alexander-Kern]

This PR adds a GitHub Actions runner for caching data from the Gin Elephant data repository and enabling local data usage for validation tests with the elephant.datasets module.
This reduces the time required for data retrieval from gin, overall reducing the runtime of the GitHub Actions runners.

GitHub Actions Runner: Added a new GitHub Actions runner to cache data from the Gin Elephant data repository. Caching action is configured to execute as a cron job on a daily basis. This automated process ensures that data caches remains up-to-date.

Local Data Usage for Validation Tests: The elephant.datasets module has been modified to use local data for validation tests. Additionally validation tests across various modules have been adapted to accommodate the option of using either local data or retrieving data from the Gin Elephant repository.
The path to local data can be provided by setting the environment variable ELEPHANT_DATA_LOCATION.

Changes to CI
Current CI runners have been adapted to retrieve the cached elephant-data.

[coveralls]

coverage: 88.401% (+0.05%) from 88.35%
when pulling f6e6353 on INM-6:enh/add_caching_action
into 0984e19 on NeuralEnsemble:master.

[Moritz-Alexander-Kern]

build wheels