Paytmlabs/

S2R_IndicVoices_hindi

Hindi speech dataset prepared from ai4bharat/IndicVoices for Ultravox training.

	Schema




	

Column Type Description

audio Audio Speech audio

text string Verbatim transcript

continuation string LLM-generated continuation (≤50 words)

Train chunks: 77/77 Validation: done