You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm reluctant to do that since size(::Tensor) already has a meaning in Julia - the Julia 'size' of any non-container object is () and it could get confusing to alter that meaning for tensors. But I am open to counter-arguments.
get_shape
seems to be TensorFlow forBase.size
.But with the ability for it to be unknown etc.
Eg
Does it makes sense to define a overload for
Base.size
, that returns a tuple ofInts
, or throws an exception if the size is unknown?The text was updated successfully, but these errors were encountered: