This is an intermediate dataset produced during the preparation of a YarnGPT-style TTS model for Asa
Combined word-aligned dataset for training Akan/Twi TTS models. Audio encoded with WavTokenizer (75H