---
license: apache-2.0
---
# Wikipedia dataset from OLLM

Repository: https://github.com/andylolu2/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`:
```python
import networkx as nx

with open(path_to_graph, "r") as f:
    G = nx.node_link_graph(json.load(f), edges="links")
```