You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When trying to synthesize speech using python: tts = TTS("tts_models/en/ljspeech/tacotron2-DDC")
It crashes python.
When running tts --text "Hello this is a test." --model_name "tts_models/en/ljspeech/tacotron2-DDC" --out_path ./test.wav
it throws the following error: zsh: bus error tts --text "Hello this is a test." --model_name --out_path ./test.wav
To Reproduce
Using Apple M1 Chip on Mac, new Python Env
brew install mecab pip install TTS
Installation is a success.
Run the following command in terminal tts --text "Hello this is a test." --model_name "tts_models/en/ljspeech/tacotron2-DDC" --out_path ./test.wav
Python crashes and no wav file is generated
Expected behavior
A TTS sound file should be successfully generated and python should not crash.
Hi! I'm not sure if this will help you, but I had a similar issue that I seem to have fixed. My machine (M2 chip) was also throwing a bus error. I found that the program crashed when initializing the AudioProcessor class. Specifically, in TTS/utils/audio/processor.py:219: self.inv_mel_basis = np.linalg.pinv(self._build_mel_basis()), the program would crash when calculating the inverse.
I fixed this by replacing np.linalg with SciPy's linalg lib, i.e.:
Describe the bug
When trying to synthesize speech using python:
tts = TTS("tts_models/en/ljspeech/tacotron2-DDC")
It crashes python.
When running
tts --text "Hello this is a test." --model_name "tts_models/en/ljspeech/tacotron2-DDC" --out_path ./test.wav
it throws the following error:
zsh: bus error tts --text "Hello this is a test." --model_name --out_path ./test.wav
To Reproduce
Using Apple M1 Chip on Mac, new Python Env
brew install mecab
pip install TTS
Installation is a success.
Run the following command in terminal
tts --text "Hello this is a test." --model_name "tts_models/en/ljspeech/tacotron2-DDC" --out_path ./test.wav
Python crashes and no wav file is generated
Expected behavior
A TTS sound file should be successfully generated and python should not crash.
Logs
Environment
Additional context
Issue #2052 was recently closed because installation on M1 chips was successfull. This was true in my case.
But the library is still not usable.
The text was updated successfully, but these errors were encountered: