!9477 Fix a bug that a and x cannot be broadcasted

From: @peixu_ren
Reviewed-by: @liangchenghui,@zichun_ye
Signed-off-by: @liangchenghui

mindspore/nn/layer/math.py

@@ -670,10 +670,10 @@ class IGamma(Cell):
         use_igammac = self.logicaland(self.greater(x, 1), self.greater(x, a))
         ax = a * self.log(x) - x - self.lgamma(a)
         para_shape = self.shape(ax)
-        boradcastto = P.BroadcastTo(para_shape)
+        broadcastto = P.BroadcastTo(para_shape)
         if para_shape != ():
-            x = boradcastto(x)
-            y = boradcastto(y)
+            x = broadcastto(x)
+            a = broadcastto(a)
         x_is_zero = self.equal(x, 0)
         log_maxfloat = self.log_maxfloat32
         underflow = self.less(ax, self.neg(log_maxfloat))
@@ -744,10 +744,10 @@ class LBeta(Cell):
         _check_input_dtype("input_y", y_dtype, x_dtype, self.cls_name)
         x_plus_y = x + y
         para_shape = self.shape(x_plus_y)
-        boradcastto = P.BroadcastTo(para_shape)
+        broadcastto = P.BroadcastTo(para_shape)
         if para_shape != ():
-            x = boradcastto(x)
-            y = boradcastto(y)
+            x = broadcastto(x)
+            y = broadcastto(y)
         comp_less = self.less(x, y)
         x_min = self.select(comp_less, x, y)
         y_max = self.select(comp_less, y, x)