napistu_torch.models.node2vec

Node2Vec utilities for Napistu-Torch.

This module provides utility functions for creating and training Node2Vec models for unsupervised node embedding learning.

Public Functions

get_node2vec_model(napistu_data, device)

Create a Node2Vec model configured for Napistu data.

get_node2vec_training_regime(model)

Get DataLoader and Optimizer for training a Node2Vec model.

get_node2vec_training_loop(model, loader, optimizer, device)

Execute a single training epoch for Node2Vec.

Functions

get_node2vec_model(napistu_data, device)
get_node2vec_training_loop(model, loader, ...)
get_node2vec_training_regime(model)

napistu_torch.models.node2vec.get_node2vec_model(napistu_data: NapistuData, device: torch.device) → torch_geometric.nn.Node2Vec

napistu_torch.models.node2vec.get_node2vec_training_loop(model: torch_geometric.nn.Node2Vec, loader: torch_geometric.data.DataLoader, optimizer: torch.optim.Optimizer, device: torch.device) → float

napistu_torch.models.node2vec.get_node2vec_training_regime(model: torch_geometric.nn.Node2Vec) → tuple[torch_geometric.data.DataLoader, torch.optim.Optimizer]