Quantifying the Performance and Resource Usage of HLS4ML's Implementation of the Batch Normalization Layer on FPGAs