In validation results, we test the performance of our pre-
diction on the same regime from which training data has
been collected. In this section, we take a leap of faith and
study the performance of our prediction on high core count
regimes (16K, 32K, and 64K). Note that since the model has
not seen points in high core count regime, it would be useful
to update them after prediction on each data point. This
adaptive approach caters to online machine learning techniques (where we update the model after seeing each point)
and is in contrast to the offine learning models used in the
validation section.