Fix for high pt EtaJES and GNNC jet calibrations

Currently, the last bin of the histograms with the spline corrections is 0. This will be fixed for future calibrations, but in the meantime, we should use the N-1 bins of the histogram to provide the extrapolation of the correction. This MR provides a fix to use that bin instead of the last one.

Tagging the conveners @mswiatlo @camacho @mleblanc @mspalla @haweber, and @dgillber who pointed out that there was a potential issue in the calibration.