| title: ColliderML Event Display | |
| emoji: 🔭 | |
| colorFrom: blue | |
| colorTo: purple | |
| sdk: gradio | |
| sdk_version: 5.50.0 | |
| python_version: "3.12" | |
| app_file: app.py | |
| pinned: false | |
| license: apache-2.0 | |
| # ColliderML Event Display | |
| Interactive 3D visualisation of single events from the | |
| [ColliderML datasets](https://huggingface.co/datasets/CERN/ColliderML-Release-1). | |
| Pick a physics process and an event ID to see the tracker hits, reconstructed | |
| tracks, and truth particles inside the OpenDataDetector geometry. | |
| ## Local development | |
| ```bash | |
| pip install -r requirements.txt | |
| python app.py | |
| # Open http://localhost:7860 | |
| ``` | |
| ## Caching | |
| The app pre-caches a small set of events per dataset on first load via | |
| `colliderml.load()`. Subsequent loads are instant. To populate the on-disk | |
| cache ahead of deploy, run: | |
| ```bash | |
| python cache_events.py | |
| ``` | |
| Cached parquets land under `_cached_events/<dataset>/*.parquet` and are | |
| picked up automatically at runtime. | |
| ## Deployment | |
| Synced automatically from `spaces/event-display/` on the | |
| [OpenDataDetector/ColliderML](https://github.com/OpenDataDetector/ColliderML) | |
| repo via the `sync-spaces.yml` workflow (see commit B7 of the v0.4.0 | |
| migration). | |