add numa to improve cpu inference perf (#2330)

Commit

1 year ago

add numa to improve cpu inference perf (#2330) Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>