Published June 16, 2022 | Version 1.0.0
Dataset Open

JetClass: A Large-Scale Dataset for Deep Learning in Jet Physics

  • 1. CERN
  • 2. Peking University

Description

JetClass is a new large-scale dataset to facilitate deep learning research in jet physics. It consists of 100M jets for training, 5M for validation and 20M for testing. The dataset contains 10 classes of jets, simulated with MadGraph + Pythia + Delphes. 

A detailed description of the JetClass dataset is presented in the paper Particle Transformer for Jet Tagging. An interface to use the dataset is provided in https://github.com/jet-universe/particle_transformer.

Files

Files (189.8 GB)

Name Size Download all
md5:64e5156d26d101adeb43b8388207d767
30.4 GB Download
md5:de4fd2dca2e68ab3c85d5cfd3bcc65c3
15.2 GB Download
md5:9722a359c5ef697bea0fbf79bf50f003
15.2 GB Download
md5:1e9f66cd1f915f9d10e90ae1d7761720
15.2 GB Download
md5:47348fc8985319fa4806da87500482fa
15.2 GB Download
md5:6b0ce16bd93b442a8d51914466990279
15.2 GB Download
md5:416e347512e716de51d392bee327b8e9
15.2 GB Download
md5:e9b9c1557b1b39bf0a16e4ab631ae451
15.2 GB Download
md5:5bfc6cb285ccb7680cefa9ac82ad1a2e
15.2 GB Download
md5:540c1a0d66dfad78d2b363c5740ccf86
15.2 GB Download
md5:668f40b3275167ff7104c48317c0ae2a
15.2 GB Download
md5:7235ccb577ed85023ea3ab4d5e6160cf
7.6 GB Download