[Bug] AttributeError: 'AttrDict' object has no attribute 'generator_model'

[dims12]

Welcome to the 🐸TTS project! We are excited to see your interest, and appreciate your support!

This repository is governed by the Contributor Covenant Code of Conduct. For more details, see the CODE_OF_CONDUCT.md file.

If you've found a bug, please provide the following information:

Describe the bug
A clear and concise description of what the bug is.

To Reproduce

1. pip install TTS
2. tts --text "Это голос дикой планеты" --model_name "tts_models/ru/ruslan/tacotron2-DDC" --vocoder_name "tts_models/ru/ruslan/tacotron2-DDC" --out_path example.wav

Downloading model to /home/dims/.local/share/tts/tts_models--ru--ruslan--tacotron2-DDC
Expected behavior
Should not crash but generate audio

Environment (please complete the following information):

$ python --version
Python 3.8.5
$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 18.04.5 LTS
Release: 18.04
Codename: bionic

• Exact command to reproduce:
  $ tts --text "Это голос дикой планеты" --model_name "tts_models/ru/ruslan/tacotron2-DDC" --vocoder_name "tts_models/ru/ruslan/tacotron2-DDC" --out_path example.wav

Downloading model to /home/dims/.local/share/tts/tts_models--ru--ruslan--tacotron2-DDC
tts_models/ru/ruslan/tacotron2-DDC is already downloaded.
Using model: Tacotron2
Traceback (most recent call last):
File "/opt/anaconda3/envs/py38/bin/tts", line 8, in
sys.exit(main())
File "/opt/anaconda3/envs/py38/lib/python3.8/site-packages/TTS/bin/synthesize.py", line 188, in main
synthesizer = Synthesizer(model_path, config_path, vocoder_path, vocoder_config_path, args.use_cuda)
File "/opt/anaconda3/envs/py38/lib/python3.8/site-packages/TTS/utils/synthesizer.py", line 49, in init
self.load_vocoder(vocoder_checkpoint, vocoder_config, use_cuda)
File "/opt/anaconda3/envs/py38/lib/python3.8/site-packages/TTS/utils/synthesizer.py", line 102, in load_vocoder
self.vocoder_model = setup_generator(self.vocoder_config)
File "/opt/anaconda3/envs/py38/lib/python3.8/site-packages/TTS/vocoder/utils/generic_utils.py", line 70, in setup_generator
print(" > Generator Model: {}".format(c.generator_model))
AttributeError: 'AttrDict' object has no attribute 'generator_model'

[erogol]

Thanks for your call @dims12 .

The problem is that you call a tts_model in place of a vocoder_model. That is --vocoder_name should take one of the vocoder we have.

Easy fix would be calling it as

$ tts --text "Это голос дикой планеты" --model_name "tts_models/ru/ruslan/tacotron2-DDC"  --out_path example.wav

then it'd use the default vocoder model set for the given tts model.

[loureirorg]

Thanks, @erogol. That works!

Maybe the instructions on README should be updated to avoid this kind of confusion, which I also ran into.

On the Run a tts and a vocoder model from the released model list. it instructs to use both --model_name and --vocoder_name at the same time.