We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
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
Describe the bug
Shape: According to the onnx reference is defined as a vector of integers
In burn_import ArgType::Shape is defined as ArgType::Shape(usize) as in it only stores the number of dimensions not the shape.
ArgType::Shape
ArgType::Shape(usize)
Expected behavior
Define ArgType::Shape(usize) as ArgType::Shape(Vec<usize>)
ArgType::Shape(Vec<usize>)
This is needed for gather op I am working on which seems to be able to accept both tensors and shapes of a tensor
Notes
If I get the time I'll work up a PR but thought I'd post here first for comments about what it is like this
The text was updated successfully, but these errors were encountered:
Looking at this close I think it would be better to make the output of the Shape convert as
ArgType::Tensor(TensorType { elem_type: ElementType::Int64, dim: 1, shape: tensor.shape.map(|shapes| vec![shapes.len()]), });
No need for a seperate ArgType as the output of Shape is a tensor of integers, which is already well described
Sorry, something went wrong.
@CohenAriel has implemented ONNX Gather, see #947.
@QuantumEntangledAndy let us know if this issue should still be open.
Closing it for now. Please re-open it if still needed.
No branches or pull requests
Describe the bug
Shape: According to the onnx reference is defined as a vector of integers
In burn_import
ArgType::Shape
is defined asArgType::Shape(usize)
as in it only stores the number of dimensions not the shape.Expected behavior
Define
ArgType::Shape(usize)
asArgType::Shape(Vec<usize>)
This is needed for gather op I am working on which seems to be able to accept both tensors and shapes of a tensor
Notes
If I get the time I'll work up a PR but thought I'd post here first for comments about what it is like this
The text was updated successfully, but these errors were encountered: