ColliderML

About the Data

The ColliderML dataset is the largest-yet source of full-detail simulation in a virtual detector experiment.

Why virtual? The simulation choices are not tied to a construction timeline, there are no budget limitations, no politics. The only goals are to produce the most realistic physics on a detailed detector geometry, with significant computating challenges, in an ML-friendly structure.

The ColliderML dataset provides comprehensive simulation data for machine learning applications in high-energy physics, with detailed detector responses and physics object reconstructions.

Get the Data

The ColliderML dataset is hosted on HuggingFace and can be accessed using the datasets library.

Prefer a more convenient workflow?

ColliderML also ships a Python library (colliderml) with a CLI download + local loading workflow and utilities (exploding tables, pileup subsampling, calibration, etc.). See Library docs.

Quick Start

1. Install the datasets library:

pip install datasets

2. Load a dataset:

from datasets import load_dataset

# Load particles data from ttbar events (no pileup)
dataset = load_dataset("CERN/Colliderml-release-1", "ttbar_pu0_particles")

Interactive Configuration

Use the configurator below to customize your dataset selection and generate the corresponding code:

Dataset Configuration

Loading available datasets from HuggingFace...

from datasets import load_dataset

# Load selected objects
particles = load_dataset("CERN/Colliderml-release-1", "ttbar_pu0_particles", split="train[:1000]")
trackerhits = load_dataset("CERN/Colliderml-release-1", "ttbar_pu0_tracker_hits", split="train[:1000]")
calohits = load_dataset("CERN/Colliderml-release-1", "ttbar_pu0_calo_hits", split="train[:1000]")
tracks = load_dataset("CERN/Colliderml-release-1", "ttbar_pu0_tracks", split="train[:1000]")

Prefer a CLI + local loader workflow? See the ColliderML library docs.

If there are errors or unexpected behavior, please open an issue on the GitHub repository.

Dataset Changelog (latest 5)

• (0.2.0 — 2025-11-07) - Datasets now hosted on HuggingFace Hub for easier access and distribution.
• (0.2.0 — 2025-11-07) - Support for standard HuggingFace datasets library for data loading.
• (0.2.0 — 2025-11-07) - Migrated from NERSC manifest-based distribution to HuggingFace datasets.
• (0.2.0 — 2025-11-07) - Data now stored in Parquet format with improved compression and accessibility.
• (0.1.0 — 2025-09-08) - Initial release of ColliderML dataset with ttbar and ggf physics processes. See the full changelog: Changelog.