Srijan-Upadhyay/

marathi-tts-indictts

	Marathi IndicTTS Speech Corpus




	


	Dataset Summary

Marathi IndicTTS Speech Corpus is a curated Marathi subset derived and formatted from IndicTTS studio recordings. It features clear, studio-recorded Marathi speech paired with phonetically balanced Devanagari transcriptions designed for acoustic feature extraction and neural vocoder training.

Format: High-fidelity WAV files (16-bit PCM, 48kHz / 22.05kHz) + normalized transcripts.… See the full description on the dataset page: https://huggingface.co/datasets/Srijan-Upadhyay/marathi-tts-indictts.