Cannot synthesize with fine-tuned model

[Ca-ressemble-a-du-fake]

Hi,

I've just got my first fine-tuned model based on Exp3 (en, pt, fr) VITS model. I want to generate a sample so I run
tts --text "{text}" \ --model_path "{modelDir}{checkpointName}" \ --config_path "{modelDir}config.json" \ --language_idx "fr-fr" \ --speaker_wav "/content/drive/MyDrive/YourTTS/TargetReferenceWavs/Test_16kHz/wavs/Test_000001.wav" \ --out_path /content/drive/MyDrive/YourTTS/output/{postfix}.wav

But instead I got
`

Traceback (most recent call last):
File "/usr/local/bin/tts", line 8, in
sys.exit(main())
File "/usr/local/lib/python3.7/dist-packages/TTS/bin/synthesize.py", line 287, in main
wav = synthesizer.tts(args.text, args.speaker_idx, args.language_idx, args.speaker_wav)
File "/usr/local/lib/python3.7/dist-packages/TTS/utils/synthesizer.py", line 323, in tts
speaker_embedding = self.tts_model.speaker_manager.compute_d_vector_from_clip(speaker_wav)
File "/usr/local/lib/python3.7/dist-packages/TTS/tts/utils/speakers.py", line 287, in compute_d_vector_from_clip
d_vector = _compute(wf)
File "/usr/local/lib/python3.7/dist-packages/TTS/tts/utils/speakers.py", line 270, in _compute
waveform = self.speaker_encoder_ap.load_wav(wav_file, sr=self.speaker_encoder_ap.sample_rate)
AttributeError: 'NoneType' object has no attribute 'load_wav'

`
I don't have speaker index to supply because I generated the dataset only with wavs folder (ljspeech like single speaker). Neither did I supply a speaker_manager in the config.json I used for training because there was an error with optimizer (can't remember which one).

So can I still test my model with the new voice I added ?

Thanks

[Ca-ressemble-a-du-fake]

I could synthesize text with my own model after providing a speaker name that I got from running tts with list_speaker_idx parameter AND (what I forgot initialy) the --config_path.

So to get speaker name :
tts --model_path "{modelDir}{checkpointName}" --config_path "{modelDir}config.json" --list_language_idxs --list_speaker_idxs

And then using that speaker name (not number even if it's called idx) :
`

!tts --text "{text}"
--model_path "{modelDir}{checkpointName}"
--config_path "{modelDir}config.json"
--language_idx "fr-fr"
--speaker_idx ljspeech
--out_path /content/drive/MyDrive/YourTTS/output/{postfix}.wav

`

Please note that ithere is no typo : speaker_idx has the same name as the dataset type that I used (ljspeech) but it could be bernard, VCTK_p225 ...