updated softmax's doc to show it now supports hessian free as well

doc/library/tensor/nnet/nnet.txt

@@ -120,11 +120,7 @@
 
    The softmax function will, when applied to a matrix, compute the softmax values row-wise.
 
-    :note: this uses a particular operation. But this method doesn't yet
-       implement the Row Operation for hessian free. If you want that, you can use
-       this equivalent code that has implemented the Row Operation
-       ``exp(x)/exp(x).sum(1, keepdims=True)``. Theano would
-       optimize this by inserting the softmax op itself.  The code of
+    :note: this supports hessian free as well.  The code of
        the softmax op is more numerically stable because it uses this code:
 
        .. code-block:: python