Unverified Commit c18f957d authored by YYY's avatar YYY Committed by GitHub
Browse files

[Model] add implementation of twirls (#2770)



* add implementation of twirls

* format the code

* fix some format error, and ignore others

* fix format errors

* fix format errors

* expose unfolding & attention

* Update nn.pytorch.rst
Co-authored-by: default avatarQuan (Andy) Gan <coin2028@hotmail.com>
parent 2a3a9a00
...@@ -136,6 +136,20 @@ DotGatConv ...@@ -136,6 +136,20 @@ DotGatConv
:members: forward :members: forward
:show-inheritance: :show-inheritance:
TWIRLSConv
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: dgl.nn.pytorch.conv.TWIRLSConv
:members: forward
:show-inheritance:
TWIRLSUnfoldingAndAttention
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: dgl.nn.pytorch.conv.TWIRLSUnfoldingAndAttention
:members: forward
:show-inheritance:
.. _apinn-pytorch-dense-conv: .. _apinn-pytorch-dense-conv:
Dense Conv Layers Dense Conv Layers
......
...@@ -21,8 +21,10 @@ from .densesageconv import DenseSAGEConv ...@@ -21,8 +21,10 @@ from .densesageconv import DenseSAGEConv
from .atomicconv import AtomicConv from .atomicconv import AtomicConv
from .cfconv import CFConv from .cfconv import CFConv
from .dotgatconv import DotGatConv from .dotgatconv import DotGatConv
from .twirlsconv import TWIRLSConv, UnfoldingAndAttention as TWIRLSUnfoldingAndAttention
__all__ = ['GraphConv', 'EdgeWeightNorm', 'GATConv', 'TAGConv', 'RelGraphConv', 'SAGEConv', __all__ = ['GraphConv', 'EdgeWeightNorm', 'GATConv', 'TAGConv', 'RelGraphConv', 'SAGEConv',
'SGConv', 'APPNPConv', 'GINConv', 'GatedGraphConv', 'GMMConv', 'SGConv', 'APPNPConv', 'GINConv', 'GatedGraphConv', 'GMMConv',
'ChebConv', 'AGNNConv', 'NNConv', 'DenseGraphConv', 'DenseSAGEConv', 'ChebConv', 'AGNNConv', 'NNConv', 'DenseGraphConv', 'DenseSAGEConv',
'DenseChebConv', 'EdgeConv', 'AtomicConv', 'CFConv', 'DotGatConv'] 'DenseChebConv', 'EdgeConv', 'AtomicConv', 'CFConv', 'DotGatConv', 'TWIRLSConv',
'TWIRLSUnfoldingAndAttention']
This diff is collapsed.
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment