Update manage.py

[Mkkrishnasai]

Issue at _set_model_item() function
While using fairseq model dataset & model variables are not set

[CLAassistant]

All committers have signed the CLA.

[eginhard]

This is already fixed in our fork (available via pip install coqui-tts). This repo is not updated anymore.

[manuel-84]

After trying this fix I get this error with TTS api

ValueError: [!] Model file not found in the output path

Code:

api = TTS("tts_models/lat/fairseq/vits")
api.tts_with_vc_to_file(
    text,
    speaker_wav="sample.wav",
    file_path="output.wav"
)

[eginhard]

@manuel-84 Can you try just running this:

api = TTS("tts_models/lat/fairseq/vits")
api.tts_to_file(
    text,
    file_path="output.wav"
)

[manuel-84]

@manuel-84 Can you try just running this:

api = TTS("tts_models/lat/fairseq/vits")
api.tts_to_file(
    text,
    file_path="output.wav"
)

Same error... I will try with your repo

[eginhard]

@manuel-84 I just checked and I can use this model without errors. If it still doesn't work with the fork, try removing the tts_models--lat--fairseq--vits/ folder, your download might have been incomplete.

[Mkkrishnasai]

@manuel-84

api = TTS("tts_models/lat/fairseq/vits")
api.tts_with_vc_to_file(
    text,
    speaker_wav="sample.wav",
    file_path="output.wav"
)

ValueError: [!] Model file not found in the output path

"api.tts_with_vc_to_file" is correct as your using "tts_models/lat/fairseq/vits" single language model with voice cloning support.

This could also be because of model download issue.

if Darwin "~/Library/Application\ Support" delete the model and try again.
different for Windows or custom path.

[rose07]

https://tts.byylook.com/ai/text-to-speech