Add occupancy hint to rms_norm

Signed-off-by: Jay Gu <jagu@nvidia.com>

Author: haijieg

test/kernels/rms_norm.py

@@ -6,7 +6,7 @@
 import numpy as np
 
 
-@ct.kernel
+@ct.kernel(occupancy=ct.ByTarget(sm_100=16))
 def rms_norm_kernel(
     x,
     w,