This dataset contains 300 decodable WAV audio files for Hausa TTS training.
audio/: Folder containing .wav audio files (24kHz)
metadata.csv: CSV with file_name, text, language, duration
metadata.jsonl: JSONL with the same metadata
Format: WAV (PCM)
Sample Rate: 24kHz
Duration: 2-15 seconds (verified)
Channels: Mono
import pandas as pd
import soundfile as sf… See the full description on the dataset page:
https://huggingface.co/datasets/vaghawan/hausa-tts-naijavoices-300.
