validation batch size uses

### UNN_BraTS23/scripts/data_loader.py
    _**line: 110 + 111**_

**Q1.** When do we use a separate batch size for validation--> Is this for validation as in nested val during training, or cross validation with external data set?

**Q2.** What is the standard validation batch size to use (i.e., same as train, half of train, relative to sample size?)