进行音频和梅尔频谱图预处理报错怎么回事： python encoder_preprocess.py <datasets_root>

The dataset consists of 0 utterances, 0 mel frames, 0 audio timesteps (0.00 hours).
Traceback (most recent call last):
File "pre.py", line 74, in
preprocess_dataset(**vars(args))
File "E:\PythonProject\MockingBird\synthesizer\preprocess.py", line 88, in preprocess_dataset
print("Max input length (text chars): %d" % max(len(m[5]) for m in metadata))
ValueError: max() arg is an empty sequence

Start with encoder_preprocess.py and pre.py, then trace the call into synthesizer/preprocess.py at line 88. Re-run the command with the dataset root and inspect why preprocessing reports zero utterances and empty metadata; done means the dataset is discovered and preprocessing completes without the shown max() error.

Written by the indexing model from the issue text.