structuration_config.json

{
  "target_repo": "Zeldeo/french-conversation_structured",
  "source": {
    "dataset_id": "Snit/french-conversation",
    "config": null,
    "split": "train",
    "columns": [
      "audio",
      "transcription",
      "id",
      "part",
      "audio_path"
    ],
    "audio_column": "audio",
    "text_column": "transcription",
    "language_column": null,
    "language_filter": "fr",
    "max_samples": 1000
  },
  "language": "fr",
  "normalization": "none",
  "add_meta": true,
  "rows_count": 98
}