Commit 7e84bf5f authored by rusty1s's avatar rusty1s
Browse files

update

parent e583d438
...@@ -15,7 +15,7 @@ As a result, our approach is provably able to maintain the expressive power of t ...@@ -15,7 +15,7 @@ As a result, our approach is provably able to maintain the expressive power of t
It provides an easy-to-use interface to convert common and custom GNNs from PyG into its scalable variant: It provides an easy-to-use interface to convert common and custom GNNs from PyG into its scalable variant:
```python ```python
from torch_geometric.nn import GCNConv from torch_geometric.nn import SAGEConv
from torch_geometric_autoscale import ScalableGNN from torch_geometric_autoscale import ScalableGNN
from torch_geometric_autoscale import metis, permute, SubgraphLoader from torch_geometric_autoscale import metis, permute, SubgraphLoader
...@@ -26,10 +26,10 @@ class GNN(ScalableGNN): ...@@ -26,10 +26,10 @@ class GNN(ScalableGNN):
pool_size=2, buffer_size=5000) pool_size=2, buffer_size=5000)
self.convs = ModuleList() self.convs = ModuleList()
self.convs.append(GCNConv(in_channels, hidden_channels)) self.convs.append(SAGEConv(in_channels, hidden_channels))
for _ in range(num_layers - 2): for _ in range(num_layers - 2):
self.convs.append(GCNConv(hidden_channels, hidden_channels)) self.convs.append(SAGEConv(hidden_channels, hidden_channels))
self.convs.append(GCNConv(hidden_channels, out_channels)) self.convs.append(SAGEConv(hidden_channels, out_channels))
def forward(self, x, adj_t, *args): def forward(self, x, adj_t, *args):
for conv, history in zip(self.convs[:-1], self.histories): for conv, history in zip(self.convs[:-1], self.histories):
......
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