File size: 3,436 Bytes
8d5f9ba
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
"""
GNN graph classifiers for Android ransomware detection.

Architectures: GIN, GCN, GAT — all using BatchNorm + global mean pooling.
Uses `PyTorchModelHubMixin` for native `from_pretrained()` / `save_pretrained()`.

Usage:
    from model import GINClassifier

    # Load from a local directory saved by save_pretrained()
    model = GINClassifier.from_pretrained("./GIN/internal_only")
    model.eval()
"""

import torch
import torch.nn as nn
from torch_geometric.nn import GINConv, GCNConv, GATConv, global_mean_pool
from huggingface_hub import PyTorchModelHubMixin


class GINClassifier(nn.Module, PyTorchModelHubMixin):
    """Graph Isomorphism Network for graph-level binary classification."""

    def __init__(self, in_dim=5, hidden=128, num_layers=3, num_classes=2, dropout=0.5):
        super().__init__()
        self.convs = nn.ModuleList()
        self.bns = nn.ModuleList()
        for i in range(num_layers):
            dim_in = in_dim if i == 0 else hidden
            mlp = nn.Sequential(
                nn.Linear(dim_in, hidden), nn.ReLU(), nn.Linear(hidden, hidden),
            )
            self.convs.append(GINConv(mlp))
            self.bns.append(nn.BatchNorm1d(hidden))
        self.classifier = nn.Sequential(
            nn.Linear(hidden, hidden), nn.ReLU(), nn.Dropout(dropout),
            nn.Linear(hidden, num_classes),
        )

    def forward(self, x, edge_index, batch):
        for conv, bn in zip(self.convs, self.bns):
            x = torch.relu(bn(conv(x, edge_index)))
        return self.classifier(global_mean_pool(x, batch))


class GCNClassifier(nn.Module, PyTorchModelHubMixin):
    """GCN baseline for graph-level classification."""

    def __init__(self, in_dim=5, hidden=128, num_layers=3, num_classes=2, dropout=0.5):
        super().__init__()
        self.convs = nn.ModuleList()
        self.bns = nn.ModuleList()
        for i in range(num_layers):
            dim_in = in_dim if i == 0 else hidden
            self.convs.append(GCNConv(dim_in, hidden))
            self.bns.append(nn.BatchNorm1d(hidden))
        self.classifier = nn.Sequential(
            nn.Linear(hidden, hidden), nn.ReLU(), nn.Dropout(dropout),
            nn.Linear(hidden, num_classes),
        )

    def forward(self, x, edge_index, batch):
        for conv, bn in zip(self.convs, self.bns):
            x = torch.relu(bn(conv(x, edge_index)))
        return self.classifier(global_mean_pool(x, batch))


class GATClassifier(nn.Module, PyTorchModelHubMixin):
    """GAT baseline for graph-level classification."""

    def __init__(self, in_dim=5, hidden=128, num_layers=3, num_classes=2, dropout=0.5, heads=4):
        super().__init__()
        self.convs = nn.ModuleList()
        self.bns = nn.ModuleList()
        for i in range(num_layers):
            dim_in = in_dim if i == 0 else hidden
            self.convs.append(GATConv(dim_in, hidden // heads, heads=heads, concat=True))
            self.bns.append(nn.BatchNorm1d(hidden))
        self.classifier = nn.Sequential(
            nn.Linear(hidden, hidden), nn.ReLU(), nn.Dropout(dropout),
            nn.Linear(hidden, num_classes),
        )

    def forward(self, x, edge_index, batch):
        for conv, bn in zip(self.convs, self.bns):
            x = torch.relu(bn(conv(x, edge_index)))
        return self.classifier(global_mean_pool(x, batch))


MODEL_REGISTRY = {"GIN": GINClassifier, "GCN": GCNClassifier, "GAT": GATClassifier}