Add XPU support to basic gnn examples(GraphSage)

examples/graph_sage_unsup.py

@@ -9,6 +9,7 @@
 from torch_geometric.datasets import Planetoid
 from torch_geometric.loader import LinkNeighborLoader
 from torch_geometric.nn import GraphSAGE
+from torch_geometric.testing.device import is_xpu_avaliable
 
 dataset = 'Cora'
 path = osp.join(osp.dirname(osp.realpath(__file__)), '..', 'data', dataset)
@@ -22,8 +23,12 @@
     neg_sampling_ratio=1.0,
     num_neighbors=[10, 10],
 )
-
-device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+if torch.cuda.is_available():
+    device = torch.device('cuda')
+elif is_xpu_avaliable():
+    device = torch.device('xpu')
+else:
+    device = torch.device('cpu')
 data = data.to(device, 'x', 'edge_index')
 
 model = GraphSAGE(

examples/graph_sage_unsup_ppi.py

@@ -12,6 +12,7 @@
 from torch_geometric.datasets import PPI
 from torch_geometric.loader import DataLoader, LinkNeighborLoader
 from torch_geometric.nn import GraphSAGE
+from torch_geometric.testing.device import is_xpu_avaliable
 
 path = osp.join(osp.dirname(osp.realpath(__file__)), '..', 'data', 'PPI')
 train_dataset = PPI(path, split='train')
@@ -29,7 +30,12 @@
 val_loader = DataLoader(val_dataset, batch_size=2)
 test_loader = DataLoader(test_dataset, batch_size=2)
 
-device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+if torch.cuda.is_available():
+    device = torch.device('cuda')
+elif is_xpu_avaliable():
+    device = torch.device('xpu')
+else:
+    device = torch.device('cpu')
 model = GraphSAGE(
     in_channels=train_dataset.num_features,
     hidden_channels=64,