Wikipedia dataset from OLLM
Train graph: train_eval_split/train_graph.json
Validation graph: train_eval_split/test_graph.json
Test graph: train_test_split/test_graph.json
Load the graph with networkx:
import networkx as nx
with open(path_to_graph, "r") as f:
G = nx.node_link_graph(json.load(f), edges="links")