Skip to content

Graph Attention Network (GAT) on Cora

Author: K3-Node Team
Backend: Multi-Backend
Dataset: Cora (Planetoid)
Description: Classify papers in the Cora citation network by topic.

View in Colab   GitHub source

Graph Attention Network (GAT) on Cora

Classify papers in the Cora citation network by topic. A GAT (Veličković et al., 2018) learns how much attention each paper should pay to each of the papers it is linked to, using 8 attention heads in the first layer.

Same model as PyG's examples/gat.py.

Install K3-Node, then choose a backend: "tensorflow", "torch" or "jax"

!pip install k3-node[examples]
import os
os.environ["KERAS_BACKEND"] = "tensorflow"

Load the data

data holds one graph: node features x, edges edge_index, labels y, and masks marking the training, validation and test nodes.

import keras
from keras import ops
from k3_node.datasets import Planetoid
from k3_node.loader import FullGraphDataset
from k3_node.transforms import NormalizeFeatures

dataset = Planetoid("data/Planetoid", name="Cora", transform=NormalizeFeatures())
data = dataset[0]
print(data)

Define the model

from k3_node.layers import GATConv


class GAT(keras.Model):
    def __init__(self, in_channels, hidden_channels, out_channels, heads):
        super().__init__()
        self.dropout = keras.layers.Dropout(0.6)
        self.conv1 = GATConv(in_channels, hidden_channels, heads=heads, dropout=0.6)
        self.conv2 = GATConv(hidden_channels * heads, out_channels, heads=1, concat=False, dropout=0.6)

    def call(self, data, training=False):
        x = self.dropout(data.x, training=training)
        x = ops.elu(self.conv1(x, data.edge_index, training=training))
        x = self.dropout(x, training=training)
        return self.conv2(x, data.edge_index, training=training)


model = GAT(dataset.num_features, hidden_channels=8, out_channels=dataset.num_classes, heads=8)

Train

FullGraphDataset feeds the whole graph to Keras; mask selects which nodes count in the loss and the accuracy.

model.compile(
    optimizer=keras.optimizers.Adam(learning_rate=0.005, weight_decay=0.0005),
    loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),
    weighted_metrics=["accuracy"],
)
model.fit(
    FullGraphDataset(data, mask="train_mask"),
    validation_data=FullGraphDataset(data, mask="val_mask"),
    epochs=200,
    verbose=2,
)

Evaluate

loss, accuracy = model.evaluate(FullGraphDataset(data, mask="test_mask"), verbose=0)
print(f"Test accuracy: {accuracy:.4f}")