Skip to content

Latest commit

 

History

History
101 lines (82 loc) · 3.34 KB

README.md

File metadata and controls

101 lines (82 loc) · 3.34 KB
Raw

Transformer-Based End-to-End Feed-Forward Neural Chinese Speech Synthesis

The text input by the client is sent to the server to synthesize speech and return, and the server and client are connected through socket to transmit data. (https://www.youtube.com/watch?v=kqExk3rd_wQ)

At present, the problem of synthesizing polyphonic words is encountered, and the International Phonetic Alphabet is used as data, which can only be judged and modified manually during synthesis.

★ Because there is a lack of part of the MelGAN vocoder code, it can only be synthesized in the Griffin Lim method. The program code is for reference only.

MelGAN vocoder is trained on the BZNSYP corpus

Installation

Make sure you have:

  • Python >= 3.6

Install espeak as phonemizer backend:

sudo apt-get install espeak-ng

Then install the rest with pip:

pip install -r requirements.txt

Custom dataset

Prepare a folder containing your metadata and wav files

|- dataset_folder/
|   |- metadata.csv
|   |- wavs/
|       |- file1.wav
|       |- ...

if metadata.csv has the following format wav_file_name|transcription you can use the bznsyp preprocessor in data/metadata_readers.py, otherwise add your own under the same file.

Make sure that:

  • the metadata reader function name is the same as data_name field in training_config.yaml.
  • the metadata file (can be anything) is specified under metadata_path in training_config.yaml

Training

Change the --config argument based on the configuration of your choice.

Train Aligner Model

Create training dataset

python create_training_data.py --config config/training_config.yaml

This will populate the training data directory (default transformer_tts_data.bznsyp).

Training

python train_aligner.py --config config/training_config.yaml

Train TTS Model

Compute alignment dataset

Use the aligner model to create the durations dataset

python extract_durations.py --config config/training_config.yaml

this will add the durations.<session name> as well as the char-wise pitch folders to the training data directory.

Training

python train_tts.py --config config/training_config.yaml

Training & Model configuration

  • Training and model settings can be configured in training_config.yaml

Resume or restart training

  • To resume training simply use the same configuration files
  • To restart training, delete the weights and/or the logs from the logs folder with the training flag --reset_dir (both) or --reset_logs, --reset_weights

Monitor training

tensorboard --logdir /logs/directory/

Prediction

In a python script

from data.audio import Audio
from model.models import ForwardTransformer
from utils.training_config_manager import TrainingConfigManager

audio = Audio.from_config(model.config)

# Feed Forward
FF_model = ForwardTransformer.load_model('/path/to/weights/')
FF_out = FF_model.predict('Please, say something.')

# Autoregressive
AR_model = config_loader.load_model()
AR_out = AR_model.predict('Please, say something.')

# Convert spectrogram to wav (with griffin lim)
FF_wav = audio.reconstruct_waveform(FF_out['mel'].numpy().T)
AR_wav = audio.reconstruct_waveform(AR_out['mel'].numpy().T)