Crate torsh_graph 

Graph Neural Network components for ToRSh

This module provides PyTorch-compatible graph neural network layers and operations, built on top of SciRS2’s graph algorithms and spectral methods.

Enhanced Features:

• GPU acceleration for graph operations
• Memory-efficient sparse representations
• Graph attention visualization
• Batch processing capabilities

Modules

attention_viz

Graph attention visualization utilities

classification

Graph-level classification networks

conv

Graph convolution layers

data

Graph data loading and manipulation utilities

datasets

Graph dataset loaders with support for popular formats

distributed

Distributed graph neural networks for large-scale graph processing

enhanced_scirs2_integration

Enhanced SciRS2 Graph Algorithm Integration

explainability

Graph explainability utilities including layer-wise relevance propagation

foundation

Graph foundation models and self-supervised learning

functional

Functional operations and activations for graph neural networks

generative

Graph Generation Models

geometric

Geometric Graph Neural Networks

hypergraph

Hypergraph Neural Networks

importance_analysis

Node importance analysis utilities

jit

Just-In-Time compilation for graph kernels

lottery_ticket

Graph Lottery Ticket Hypothesis

matching

Graph Matching and Similarity Learning

multimodal

Multi-Modal Graph Learning

neural_operators

Graph Neural Operators

neuromorphic

Neuromorphic graph processing - Bio-inspired graph neural networks

optimal_transport

Graph Optimal Transport

parameter

Minimal Parameter implementation for torsh-graph This is a temporary replacement for torsh_nn::Parameter to avoid compilation issues

pool

Graph pooling layers

quantum

Quantum graph algorithms and quantum-inspired graph neural networks

scirs2_integration

SciRS2 Graph Integration

spectral

Spectral Graph Neural Networks

temporal

Temporal Graph Neural Networks

utils

Graph utilities and algorithms

Structs

GraphData

Graph data structure for GNNs

GraphMemoryStats

Memory usage statistics for a graph

Enums

GraphValidationError

Graph validation errors

Traits

GraphLayer

Trait for graph neural network layers