-
-
Notifications
You must be signed in to change notification settings - Fork 152
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Continue cleaning up the sparse/dense tensor types #886
Comments
Sounds less painful to keep the current |
I agree, and although |
Yeah, we can make |
Hi, I can work on this task. Thanks |
That would be great! |
Currently,
TensorType
s aren't an abstract class, and direct instances ofTensorType
s represent dense tensors, as do instances of the newDenseTensorType
.Instead, it would be better to have and use a single abstract tensor
Type
and distinct sparse and dense tensorType
s. In this case, we could remove all the meta classes.One of the issues I can see with this change: external libraries that use
TensorType
and/orTensorVariable
s directly will need to change these toDenseTensorType
andDenseTensorVariable
, unless we continue to keep these names reserved for dense tensors and use newAbstractTensor[Type|Variable]
classes as the shared base.Originally posted by @brandonwillard in #766 (review)
The text was updated successfully, but these errors were encountered: