This is the repository for creating the dashboard at https://hydrogen-event-tree.github.io/ where we present a configurable event tree based on events from the The Hydrogen Incident and Accidents Database HIAD 2.1, developed by The Joint Research Centre (JRC) of the European Commission. This project was developed as a part of HYDROGENi, a Norwegian research and innovation centre for hydrogen and ammonia. More details are available in the paper (link coming soon) which this dashboard is a part of.
Obviously, we provide no guarantees about the correctness of any of the information on the dashboard but rather present it as our best effort at creating a useful event tree from the events in HIAD. Feel free to submit PRs or open issues if you feel anything should be changed, or fork a version yourself.
The dashboard loads events-manifest.json in the same top-level folder as index.html. The manifest lists one or more model entries:
{
"models": [
{
"id": "qwen3-14b",
"name": "Qwen3-14B",
"events_path": "events-qwen3-14b.json",
"model": "qwen3:14b"
}
],
"default_model_id": "qwen3-14b"
}Each events_path points to a JSON array produced by main.py. To generate (or refresh) these files, set gen = 1 in main.py. It will run process_events() for every entry in LLMS, write events-<id>.json files, and rebuild events-manifest.json. If you already have the JSON files, leave gen = 0 and the script will just start the webserver on localhost:4000.
Models are configured in LLMS at the top of main.py. For local inference, ollama must be installed both as a pip package and as a running service (https://ollama.com/). For OpenRouter-backed entries, set OPENROUTER_API_KEY in main.py before running the generator, and ensure the openai package is installed.