How to update part of a shared variable in Theano?
E.g. instead of doing:
gradient_W = T.grad(cost,W)
updates.append((W, W-learning_rate*gradient_W))
train = th.function(inputs=[index], outputs=[cost], updates=updates,
givens={x:self.X[index:index+mini_batch_size,:]})
I'd like to only update part of W, e.g. only the first column:
updates.append((W[:, 0], W[:, 0]-learning_rate*gradient_W))
But doing so gives the error TypeError: ('update target must be a SharedVariable', Subtensor{::, int64}.0):
Traceback (most recent call last):
File "ae.py", line 330, in <module>
main()
File "ae.py", line 305, in main
ae.train(n_epochs=n_epochs, mini_batch_size=100, learning_rate=0.002, train_data= train_sentence_embeddings, test_data= test_sentence_embeddings)
File "ae.py", line 87, in train
givens={x:self.X[index:index+mini_batch_size,:]})
File "/usr/local/lib/python2.7/dist-packages/theano/compile/function.py", line 266, in function
profile=profile)
File "/usr/local/lib/python2.7/dist-packages/theano/compile/pfunc.py", line 489, in pfunc
no_default_updates=no_default_updates)
File "/usr/local/lib/python2.7/dist-packages/theano/compile/pfunc.py", line 194, in rebuild_collect_shared
store_into)
TypeError: ('update target must be a SharedVariable', Subtensor{::, int64}.0)
What is the typical way to do so in Theano?
W is a typical weight matrix:
initial_W = np.asarray(rng.uniform(
low=-4 * np.sqrt(6. / (self.hidden_size + self.n)),
high=4 * np.sqrt(6. / (self.hidden_size + self.n)),
size=(self.n, self.hidden_size)), dtype=th.config.floatX)
W = th.shared(value=initial_W, name='W', borrow=True)
You can use e.g. theano.tensor.set_subtensor.
Taking the updates line you mention above, this becomes:
updates.append((W, T.set_subtensor(W[:, 0], W[:, 0]-learning_rate*gradient_W)))
where T = theano.tensor.
Another possibility, a little more concise in your case, is T.inc_subtensor, where you specify only the increment instead of the new value:
updates.append((W, T.inc_subtensor(W[:, 0], -learning_rate*gradient_W)))