Build and train graph neural networks with PyTorch.
npx skills add https://github.com/galeep/plugin-place --skill torch-geometric-galeep
Please help me install this Agent Skill. Skill: torch-geometric Source: https://github.com/galeep/plugin-place/tree/main/plugins/sci-machine-learning/skills/torch-geometric Command: npx skills add https://github.com/galeep/plugin-place --skill torch-geometric-galeep
This Skill provides a comprehensive set of tools for building and training graph neural networks (GNNs) using PyTorch, addressing the challenges of working with graph-structured data.
Use the torch-geometric skill to build a simple GCN model for node classification:
import torch
import torch.nn.functional as F
from torch_geometric.nn import GCNConv
class GCN(torch.nn.Module):
def __init__(self, in_channels, hidden_channels, out_channels):
super().__init__()
self.conv1 = GCNConv(in_channels, hidden_channels)
self.conv2 = GCNConv(hidden_channels, out_channels)
def forward(self, x, edge_index):
x = self.conv1(x, edge_index).relu()
x = F.dropout(x, p=0.5, training=self.training)
x = self.conv2(x, edge_index)
return x
High-intent search queries and answers about installing and using this skill.
To build a graph neural network for node classification with PyTorch, use built-in GNN layers like GCNConv to process graph-structured data and generate node-level predictions. You can stack multiple layers to extract graph features.
Yes, you can train graph neural networks on large graphs using mini-batch strategies. This approach scales GNN training by processing subsets of nodes and edges, making it feasible to handle massive graph-structured data efficiently.
Yes, this PyTorch graph neural network skill supports heterogeneous graphs. It provides built-in data structures to easily handle graphs containing multiple types of nodes and edges within the same network structure.
The best way to perform link prediction on graph-structured data in PyTorch is by using specialized GNN layers to learn node embeddings. These embeddings capture structural relationships to accurately predict missing or future edges.
Yes, you need PyTorch 2.6 or higher and torch-geometric 2.7.x or higher to use GNN layers like GAT and SAGE. The environment also requires Python 3.10 or above to run properly.
When scaling graph classification models, limitations arise from memory constraints when loading entire large graphs. To overcome this, use mini-batch training strategies to partition graph-structured data into manageable batches.
tassiovaleApply graph neural networks using the PyTorch Geometric library.
crazymsnBuild and evaluate graph neural networks with PyG for node classification, link prediction, and heterogeneous graphs.
viniruggeriBuild and optimize Graph Neural Network models with PyG for node classification, link prediction, and graph classification.
Hawksight-AIGenerate node2vec embeddings for similarity search and link prediction in graphs.