chainer.functions.forget(func, *xs)[source]

Calls a function without storing intermediate results.

On a forward propagation, Chainer normally stores all intermediate results of VariableNodes on a computational graph as they are required on backward propagation. Sometimes these results consume too much memory. F.forget forgets such intermediate results on forward propagation, and still supports backpropagation with recalculation.

On a forward propagation, F.forget calls a given function with given variables without creating a computational graph. That means, no intermediate results are stored. On a backward propagation, F.forget calls the given function again to create a computational graph for backpropagation.

F.forget reduces internal memory usage, whereas it requires more calculation time as it calls the function twice.


Let f be a function defined as:

>>> def f(a, b):
...   return (a + b) * a

and, x and y be Variables:

>>> x = chainer.Variable(np.random.uniform(-1, 1, 5).astype(np.float32))
>>> y = chainer.Variable(np.random.uniform(-1, 1, 5).astype(np.float32))

When z is calculated as z = f(x, y), its intermediate result x + y is stored in memory. Instead, if you call f with F.forget:

>>> z = F.forget(f, x, y)

intermediate x + y is forgotten.


F.forget does not support functions which behave differently in multiple calls with the same inputs, such as F.dropout() and F.negative_sampling().


In case input argument variables are of N-dimensional array objects, arguments will automatically be converted to Variables. This conversion takes place to ensure that this function is included in the computational graph to enable backward computations.


F.forget does not support double backpropagation.


If you want to use F.forget to a link which updates the link’s internal information every time the forward computation is called, please ensure that the information is updated just once in a single iteration. You may use the chainer.config.in_recomputing flag to check if the forward computation is the first call in an iteration. Please see the implementation of BatchNormalization for detail.


A variable func returns. If it returns a tuple, the method returns a tuple too.

Return type