stats/data_scope.json

{
  "description": "Curated, balanced real and synthetic five-language speech",
  "languages": [
    "en",
    "de",
    "fr",
    "nl",
    "es"
  ],
  "source_fractions": {
    "real": 0.5,
    "synthetic": 0.5
  },
  "redistributed_inputs": false,
  "provider_identifying_metadata_omitted": true,
  "raw_audio_redistributed": false,
  "byte_identical_data_reproduction_claimed": false
}