{
  "READY": true,
  "name": "QM8_f1-CC2",
  "task": "regression",
  "cat_idx": [],
  "num_samples": 21747,
  "num_features": 676,
  "description": "Electronic spectra and excited state energy of small molecules calculated by multiple quantum mechanic methods. Task: predict the 'f1-CC2' label using Coulomb matrix features. Target: z-scored with the training-split mean and SD (stored as target_mean / target_std), following DeepChem's default target normalization for this loader. Split: random 70/15/15 (seed 42).",
  "source": "MoleculeNet (https://moleculenet.org) \u2014 loaded via DeepChem.",
  "label": "Drug Discovery",
  "sub_labels": [
    "Quantum Mechanics",
    "Coulomb Matrix"
  ]
}