[Bug] ValueError: Model is not multi-lingual but language is provided.

[JJun-Guo]

Describe the bug

I got the "ValueError: Model is not multi-lingual but language is provided." when i used XTTSv2

To Reproduce

import torch
from TTS.api import TTS

Get device

device = "cuda" if torch.cuda.is_available() else "cpu"

List available 🐸TTS models

print(TTS().list_models())

Init TTS

tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2").to(device)

Run TTS

❗ Since this model is multi-lingual voice cloning model, we must set the target speaker_wav and language

Text to speech list of amplitude values as output

wav = tts.tts(text="Hello world!", speaker_wav="my/cloning/audio.wav", language="en")

Text to speech to a file

tts.tts_to_file(text="Hello world!", speaker_wav="my/cloning/audio.wav", language="en", file_path="output.wav")

Expected behavior

No response

Logs

No response

Environment

Environment u provided

Additional context

No response

[jackyin68]

I just commented out parameter check line, and waiting for fix.

[ANK789]

em, the same question,waiting

[yhyu13]

I encountered the same error msg when loading from local model checkpoint,

turns out you need to input the model name for xtts to be recognized as multilingual.

But since you guys load from model name instead of loading from a given local model checkpoint, xtts2 should be fine with multilingual, this is strange

[Edresson]

Hi @JJun-Guo @jackyin68 @yhyu13 @ANK789,

It was fixed on #3273. Please try to update your 🐸 TTS version to the latest one :).

[JJun-Guo]

Hi @JJun-Guo @jackyin68 @yhyu13 @ANK789,

It was fixed on #3273. Please try to update your 🐸 TTS version to the latest one :).

But i got the other bug " File "./TTS/TTS/tts/layers/xtts/tokenizer.py", line 653, in encode
return self.tokenizer.encode(txt).ids
AttributeError: 'NoneType' object has no attribute 'encode'"

I have done !!!