This is a comprehensive Urdu Text-to-Speech (TTS) dataset featuring 17,601 couplets (ash'aar) from 15 legendary Urdu poets. The audio was synthesized using the ur-PK-AsadNeural voice, specifically optimized for poetic recitation with custom pauses and a rhythmic pace (-25% rate).

📊 Dataset Highlights

Total Examples: 17,601 Total Poets: 15 Format: Parquet (Embedded Audio) Audio Quality: 24kHz Mono Size: ~723 MB… See the full description on the dataset page: https://huggingface.co/datasets/Khurram123/Urdu-Poetry-TTS-Dataset.