rhasspy
Error when training model
Hi,
I get this weird error when training a model on my Debian server using a P4 GPU using this command python3 -m piper_train.preprocess --language en --input-dir /home/ishaa/AI/dataprep --output-dir /home/ishaa/AI/training/train-me --dataset-format ljspeech --single-speaker --sample-rate 22050
The error I get with what came before it I am new please let me know if this is a issue that has been had before or if i am missing something obvious I am using this guide to train:https://blog.networkchuck.com/posts/how-to-clone-a-voice/
`INFO:preprocess:Single speaker dataset INFO:preprocess:Wrote dataset config INFO:preprocess:Processing 179 utterance(s) with 6 worker(s) /home/ishaa/AI/training/piper/src/python/piper_train/norm_audio/init.py:52: UserWarning: PySoundFile failed. Trying audioread instead. audio_16khz, _sr = librosa.load(path=audio_path, sr=vad_sample_rate) /home/ishaa/AI/training/.venv/lib/python3.11/site-packages/librosa/core/audio.py:184: FutureWarning: librosa.core.audio.__audioread_load Deprecated as of librosa version 0.10.0. It will be removed in librosa version 1.0. y, sr_native = __audioread_load(path, offset, duration, dtype) ERROR:preprocess:Failed to process utterance: Utterance(text='ago, RTX formerly Rayfion was fined $950 million for inflating bills to the Department of Defense as they lied.', audio_path=PosixPath('/home/ishaa/AI/dataprep/wav/split_Bernie1_nosilence_039.wav'), speaker=None, speaker_id=None, phonemes=['ɐ', 'ɡ', 'ˈ', 'ə', 'ʊ', ',', ' ', 'ˌ', 'ɑ', 'ː', 't', 'ˌ', 'i', 'ː', 'ˈ', 'ɛ', 'k', 's', ' ', 'f', 'ˈ', 'ɔ', 'ː', 'm', 'ə', 'l', 'i', ' ', 'ɹ', 'ˈ', 'e', 'ɪ', 'f', 'i', 'ə', 'n', ' ', 'w', 'ɒ', 'z', ' ', 'f', 'ˈ', 'a', 'ɪ', 'n', 'd', ' ', 'd', 'ˈ', 'ɒ', 'l', 'ə', ' ', 'n', 'ˈ', 'a', 'ɪ', 'n', 'h', 'ˈ', 'ʌ', 'n', 'd', 'ɹ', 'ɪ', 'd', 'ə', 'n', ' ', 'f', 'ˈ', 'ɪ', 'f', 't', 'i', ' ', 'm', 'ˈ', 'ɪ', 'l', 'i', 'ə', 'n', ' ', 'f', 'ɔ', 'ː', 'ɹ', ' ', 'ɪ', 'n', 'f', 'l', 'ˈ', 'e', 'ɪ', 't', 'ɪ', 'ŋ', ' ', 'b', 'ˈ', 'ɪ', 'l', 'z', ' ', 't', 'ə', ' ', 'ð', 'ə', ' ', 'd', 'ɪ', 'p', 'ˈ', 'ɑ', 'ː', 't', 'm', 'ə', 'n', 't', ' ', 'ɒ', 'v', ' ', 'd', 'ɪ', 'f', 'ˈ', 'ɛ', 'n', 's', ' ', 'a', 'z', ' ', 'ð', 'e', 'ɪ', ' ', 'l', 'ˈ', 'a', 'ɪ', 'd', '.'], phoneme_ids=[1, 0, 50, 0, 66, 0, 120, 0, 59, 0, 100, 0, 8, 0, 3, 0, 121, 0, 51, 0, 122, 0, 32, 0, 121, 0, 21, 0, 122, 0, 120, 0, 61, 0, 23, 0, 31, 0, 3, 0, 19, 0, 120, 0, 54, 0, 122, 0, 25, 0, 59, 0, 24, 0, 21, 0, 3, 0, 88, 0, 120, 0, 18, 0, 74, 0, 19, 0, 21, 0, 59, 0, 26, 0, 3, 0, 35, 0, 52, 0, 38, 0, 3, 0, 19, 0, 120, 0, 14, 0, 74, 0, 26, 0, 17, 0, 3, 0, 17, 0, 120, 0, 52, 0, 24, 0, 59, 0, 3, 0, 26, 0, 120, 0, 14, 0, 74, 0, 26, 0, 20, 0, 120, 0, 102, 0, 26, 0, 17, 0, 88, 0, 74, 0, 17, 0, 59, 0, 26, 0, 3, 0, 19, 0, 120, 0, 74, 0, 19, 0, 32, 0, 21, 0, 3, 0, 25, 0, 120, 0, 74, 0, 24, 0, 21, 0, 59, 0, 26, 0, 3, 0, 19, 0, 54, 0, 122, 0, 88, 0, 3, 0, 74, 0, 26, 0, 19, 0, 24, 0, 120, 0, 18, 0, 74, 0, 32, 0, 74, 0, 44, 0, 3, 0, 15, 0, 120, 0, 74, 0, 24, 0, 38, 0, 3, 0, 32, 0, 59, 0, 3, 0, 41, 0, 59, 0, 3, 0, 17, 0, 74, 0, 28, 0, 120, 0, 51, 0, 122, 0, 32, 0, 25, 0, 59, 0, 26, 0, 32, 0, 3, 0, 52, 0, 34, 0, 3, 0, 17, 0, 74, 0, 19, 0, 120, 0, 61, 0, 26, 0, 31, 0, 3, 0, 14, 0, 38, 0, 3, 0, 41, 0, 18, 0, 74, 0, 3, 0, 24, 0, 120, 0, 14, 0, 74, 0, 17, 0, 10, 0, 2], audio_norm_path=None, audio_spec_path=None, missing_phonemes=Counter()) Traceback (most recent call last): File "/home/ishaa/AI/training/.venv/lib/python3.11/site-packages/librosa/core/audio.py", line 176, in load y, sr_native = __soundfile_load(path, offset, duration, dtype) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ishaa/AI/training/.venv/lib/python3.11/site-packages/librosa/core/audio.py", line 209, in __soundfile_load context = sf.SoundFile(path) ^^^^^^^^^^^^^^^^^^ File "/home/ishaa/AI/training/.venv/lib/python3.11/site-packages/soundfile.py", line 690, in init self._file = self._open(file, mode_int, closefd) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ishaa/AI/training/.venv/lib/python3.11/site-packages/soundfile.py", line 1265, in _open raise LibsndfileError(err, prefix="Error opening {0!r}: ".format(self.name)) soundfile.LibsndfileError: Error opening '/home/ishaa/AI/dataprep/wav/split_Bernie1_nosilence_039.wav': Format not recognised.
During handling of the above exception, another exception occurred:
Traceback (most recent call last): File "/home/ishaa/AI/training/piper/src/python/piper_train/preprocess.py", line 315, in phonemize_batch_espeak utt.audio_norm_path, utt.audio_spec_path = cache_norm_audio( ^^^^^^^^^^^^^^^^^ File "/home/ishaa/AI/training/piper/src/python/piper_train/norm_audio/init.py", line 52, in cache_norm_audio audio_16khz, _sr = librosa.load(path=audio_path, sr=vad_sample_rate) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ishaa/AI/training/.venv/lib/python3.11/site-packages/librosa/core/audio.py", line 184, in load y, sr_native = __audioread_load(path, offset, duration, dtype) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ishaa/AI/training/.venv/lib/python3.11/site-packages/decorator.py", line 232, in fun return caller(func, *(extras + args), **kw) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ishaa/AI/training/.venv/lib/python3.11/site-packages/librosa/util/decorators.py", line 59, in __wrapper return func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File "/home/ishaa/AI/training/.venv/lib/python3.11/site-packages/librosa/core/audio.py", line 240, in __audioread_load reader = audioread.audio_open(path) ^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ishaa/AI/training/.venv/lib/python3.11/site-packages/audioread/init.py", line 132, in audio_open raise NoBackendError() audioread.exceptions.NoBackendError`
