rtxtd/tts-dataset-guj-eng

Indian TTS Dataset

A high-quality Text-to-Speech training dataset of 60 minutes total duration:

• 30 minutes Indian English (en-IN)
• 30 minutes Gujarati (gu-IN)

Dataset Details

Field	Value
Total samples	120
Total duration	60 minutes
en-IN samples	60 (30 sec each)
gu-IN samples	60 (30 sec each)

Schema

Each sample contains:

{{

    "audio":              AudioFile,      # WAV, 16kHz mono
    "language_code":      str,            # "en-IN" or "gu-IN"
    "speaker_id":         str,            # e.g. "eng_spk_01"
    "duration":           float,          # seconds
    "transcript":         str,            # cleaned text
    "emotion":            str,            # emotion label
    "emotion_confidence": float,          # 0–1
    "style":              str,            # speaking style
    "segment_id":         str,
}}