-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathmodels.py
211 lines (161 loc) · 7.75 KB
/
models.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
import dgl
import torch
import torch.nn as nn
import torch.nn.functional as F
from dgl.nn import GraphConv, GATConv, SAGEConv, APPNPConv
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
class MLP(nn.Module):
def __init__(self, num_layers, input_dim, hidden_dim, output_dim, dropout_ratio):
super(MLP, self).__init__()
self.num_layers = num_layers
self.dropout = nn.Dropout(dropout_ratio)
self.layers = nn.ModuleList()
if num_layers == 1:
self.layers.append(nn.Linear(input_dim, output_dim))
else:
self.layers.append(nn.Linear(input_dim, hidden_dim))
for i in range(num_layers - 2):
self.layers.append(nn.Linear(hidden_dim, hidden_dim))
self.layers.append(nn.Linear(hidden_dim, output_dim))
def forward(self, feats):
h = feats
h_list = []
for l, layer in enumerate(self.layers):
h = layer(h)
if l != self.num_layers - 1:
h = F.relu(h)
h = self.dropout(h)
h_list.append(h)
return h_list, h
class GCN(nn.Module):
def __init__(self, num_layers, input_dim, hidden_dim, output_dim, dropout_ratio, activation):
super(GCN, self).__init__()
self.num_layers = num_layers
self.dropout = nn.Dropout(dropout_ratio)
self.layers = nn.ModuleList()
if num_layers == 1:
self.layers.append(GraphConv(input_dim, output_dim, activation=activation))
else:
self.layers.append(GraphConv(input_dim, hidden_dim, activation=activation))
for i in range(num_layers - 2):
self.layers.append(GraphConv(hidden_dim, hidden_dim, activation=activation))
self.layers.append(GraphConv(hidden_dim, output_dim))
def forward(self, g, feats):
h = feats
h_list = []
for l, layer in enumerate(self.layers):
h = layer(g, h)
if l != self.num_layers - 1:
h = self.dropout(h)
h_list.append(h)
return h_list, h
class GAT(nn.Module):
def __init__(
self,num_layers, input_dim, hidden_dim, output_dim, dropout_ratio, activation, num_heads=8, attn_drop=0.3, negative_slope=0.2, residual=False):
super(GAT, self).__init__()
hidden_dim //= num_heads
self.num_layers = num_layers
self.layers = nn.ModuleList()
heads = ([num_heads] * num_layers) + [1]
self.layers.append(GATConv(input_dim, hidden_dim, heads[0], dropout_ratio, attn_drop, negative_slope, False, activation))
for l in range(1, num_layers - 1):
self.layers.append(GATConv(hidden_dim * heads[l-1], hidden_dim, heads[l], dropout_ratio, attn_drop, negative_slope, residual, activation))
self.layers.append(GATConv(hidden_dim * heads[-2], output_dim, heads[-1], dropout_ratio, attn_drop, negative_slope, residual, None))
def forward(self, g, feats):
h = feats
h_list = []
for l, layer in enumerate(self.layers):
h = layer(g, h)
if l != self.num_layers - 1:
h = h.flatten(1)
else:
h = h.mean(1)
h_list.append(h)
return h_list, h
class APPNP(nn.Module):
def __init__(self, num_layers, input_dim, hidden_dim, output_dim, dropout_ratio, activation, edge_drop=0.5, alpha=0.1, k=10):
super(APPNP, self).__init__()
self.num_layers = num_layers
self.activation = activation
self.layers = nn.ModuleList()
self.dropout = nn.Dropout(dropout_ratio)
if num_layers == 1:
self.layers.append(nn.Linear(input_dim, output_dim))
else:
self.layers.append(nn.Linear(input_dim, hidden_dim))
for i in range(num_layers - 2):
self.layers.append(nn.Linear(hidden_dim, hidden_dim))
self.layers.append(nn.Linear(hidden_dim, output_dim))
self.propagate = APPNPConv(k, alpha, edge_drop)
self.reset_parameters()
def reset_parameters(self):
for layer in self.layers:
layer.reset_parameters()
def forward(self, g, feats):
h = feats
h_list = []
h = self.activation(self.layers[0](self.dropout(h)))
for l, layer in enumerate(self.layers[1:-1]):
h = self.activation(layer(h))
h = self.layers[-1](self.dropout(h))
h = self.propagate(g, h)
h_list.append(h)
return h_list, h
class GraphSAGE(nn.Module):
def __init__(self, num_layers, input_dim, hidden_dim, output_dim, dropout_ratio, activation):
super(GraphSAGE, self).__init__()
self.num_layers = num_layers
self.dropout = nn.Dropout(dropout_ratio)
self.layers = nn.ModuleList()
self.activation = activation
if num_layers == 1:
self.layers.append(SAGEConv(input_dim, output_dim, aggregator_type='gcn'))
else:
self.layers.append(SAGEConv(input_dim, hidden_dim, aggregator_type='gcn'))
for i in range(num_layers - 2):
self.layers.append(SAGEConv(hidden_dim, hidden_dim, aggregator_type='gcn'))
self.layers.append(SAGEConv(hidden_dim, output_dim, aggregator_type='gcn'))
def forward(self, g, feats):
h = feats
h_list = []
for l, layer in enumerate(self.layers):
h = layer(g, h)
if l != self.num_layers - 1:
h = self.activation(h)
h = self.dropout(h)
h_list.append(h)
return h_list, h
# Implementation for teacher GNNs and student MLPs
class Model(nn.Module):
def __init__(self, param):
super(Model, self).__init__()
if param['distill_mode'] == 0:
self.model_name = param["teacher"]
else:
self.model_name = param["student"]
if "MLP" in self.model_name:
self.encoder = MLP(num_layers=param["num_layers"], input_dim=param["feat_dim"], hidden_dim=param["hidden_dim"], output_dim=param["label_dim"], dropout_ratio=param["dropout_t"]).to(device)
elif "GCN" in self.model_name:
self.encoder = GCN(num_layers=param["num_layers"], input_dim=param["feat_dim"], hidden_dim=param["hidden_dim"], output_dim=param["label_dim"], dropout_ratio=param["dropout_s"], activation=F.relu).to(device)
elif "GAT" in self.model_name:
self.encoder = GAT(num_layers=param["num_layers"], input_dim=param["feat_dim"], hidden_dim=param["hidden_dim"], output_dim=param["label_dim"], dropout_ratio=param["dropout_s"], activation=F.relu, num_heads=param['num_heads']).to(device)
elif "SAGE" in self.model_name:
self.encoder = GraphSAGE(num_layers=param["num_layers"], input_dim=param["feat_dim"], hidden_dim=param["hidden_dim"], output_dim=param["label_dim"], dropout_ratio=param["dropout_s"], activation=F.relu).to(device)
elif "APPNP" in self.model_name:
self.encoder = APPNP(num_layers=param["num_layers"], input_dim=param["feat_dim"], hidden_dim=param["hidden_dim"], output_dim=param["label_dim"], dropout_ratio=param["dropout_s"], activation=F.relu).to(device)
def forward(self, data, feats):
if "MLP" in self.model_name:
return self.encoder(feats)[1]
else:
return self.encoder(data, feats)[1]
def edge_distribution_high(self, edge_idx, feats, tau):
src = edge_idx[1][0]
dst = edge_idx[1][1]
feats_abs = torch.abs(feats[src] - feats[dst])
e_softmax = F.log_softmax(feats_abs / tau, dim=-1)
return e_softmax
def edge_distribution_low(self, edge_idx, feats, out, criterion_t):
src = edge_idx[0][0]
dst = edge_idx[0][1]
loss = criterion_t(feats[src], out[dst])
return loss