Update Option of Theano Function -

i using theano tutorials of deep learning , have doubt regarding how theano function's update works. whatever parameters update defined take new value if of parameters changed?

theano.function( [i],                  cost,                  updates = updates,                  givens = { self.x: trainx[i*self.mbsize: (i+1)*self.mbsize],                             self.y: trainy[i*self.mbsize: (i+1)*self.mbsize]                           }                ) 

updates defined

updates = [ (param, param-learnrate*grad)                     param, grad in zip(self.params, gradients) ] 

here learnrate not theano variable. question if change learnrate @ point of time theano function take changed value of learnrate or continue old value?

learnrate changed as

learnrate = learnrate/10 

initially learnrate 0.05

it continue old value.

numbers immutable in python changing value of learnrate won't affect value being used in updates list, , guess once theano function compiled provided updates can not changed unless compile again new value.

as have known, can make learning rate theano variable , change value when want to.

here's question might useful, is there way change function's update list without re-compiling in theano?