config.json

{
  "version": "1.0.0",
  "description": "Synthetic Pashto speech dataset with 5000 sentences in male and female voices (Parquet format)",
  "citation": "@dataset{pashto_synthetic_speech_parquet_5k, title = {Pashto Synthetic Speech Dataset Parquet (5k)}, year = {2025}}",
  "homepage": "https://huggingface.co/datasets/username/pashto_speech_5k",
  "license": "mit",
  "features": {
    "sentence_id": {
      "dtype": "string",
      "id": null,
      "_type": "Value"
    },
    "sentence": {
      "dtype": "string",
      "id": null,
      "_type": "Value"
    },
    "audio": {
      "_type": "Audio",
      "dtype": "struct",
      "id": null,
      "sampling_rate": 24000
    },
    "duration": {
      "dtype": "float64",
      "id": null,
      "_type": "Value"
    },
    "speaker": {
      "dtype": "string",
      "id": null,
      "_type": "Value"
    },
    "sampling_rate": {
      "dtype": "int64",
      "id": null,
      "_type": "Value"
    }
  }
}