from pathlib import Path

import soundfile as sf

from datasets import load_dataset

dataset = load_dataset("mythicinfinity/libritts", "clean")
for split in dataset.keys():
    Path(f"libritts/{split}").mkdir(exist_ok=True)
    for i, x in enumerate(dataset[split]):
        # audio = x["audio"]
        text = x["text_normalized"]
        # wav, sr = audio["array"], audio["sampling_rate"]
        # sf.write(f"libritts/{split}/{i}.wav", wav, sr)
        with open(f"libritts/{split}/{i}.txt", "w") as f:
            f.write(text)