Upload dataset_info.json with huggingface_hub

dataset_info.json

@@ -0,0 +1,28 @@
+{
+  "name": "Pashto Synthetic Speech Dataset Parquet (20k)",
+  "version": "1.0.0",
+  "description": "Synthetic Pashto speech dataset with 20000 sentences in male and female voices (Parquet format)",
+  "license": "MIT",
+  "created_at": "2025-04-24T10:38:23.434064",
+  "format": "parquet",
+  "statistics": {
+    "total_sentences": 20000,
+    "total_male_recordings": 20000,
+    "total_female_recordings": 20000,
+    "total_recordings": 40000,
+    "sampling_rate": 24000,
+    "total_male_duration_seconds": 19979.763,
+    "total_female_duration_seconds": 20440.578,
+    "average_male_duration_seconds": 0.99898815,
+    "average_female_duration_seconds": 1.0220289,
+    "longest_male_clip_seconds": 2.115,
+    "longest_female_clip_seconds": 2.196,
+    "shortest_male_clip_seconds": 0.228,
+    "shortest_female_clip_seconds": 0.228
+  },
+  "voices": {
+    "male": "ps-AF-GulNawazNeural",
+    "female": "ps-AF-LatifaNeural"
+  },
+  "audio_format": "WAV, 24kHz, 16-bit PCM"
+}