Make specifying dataset splits optional

[hperrot]

Is your feature request related to a problem? Please describe.
Sometimes only one datset ist required. The implementation in the dev branch however requires a separate definition of a train and validation dataset.

Describe the solution you'd like
Maybe if only one dataset is given, if edflow is started with -t, this should be taken as train dataset else the validation dataset. Then only this split should be evelauated. In case of a Training, the Evaluation after the epochs should also not be done.

Describe alternatives you've considered
For now I specify the same dataset for both splits in such a case.