ilovejs / transformertts Goto Github PK

Implementation of a non-autoregressive Transformer based neural network for Text-to-Speech (TTS).
This repo is based on the following papers:

• Neural Speech Synthesis with Transformer Network
• FastSpeech: Fast, Robust and Controllable Text to Speech

Spectrograms produced with LJSpeech and standard data configuration from this repo are compatible with WaveRNN.

Being non-autoregressive, this Transformer model is:

• Robust: No repeats and failed attention modes for challenging sentences.
• Fast: With no autoregression, predictions take a fraction of the time.
• Controllable: It is possible to control the speed of the generated utterance.

Can be found here.

These samples' spectrograms are converted using the pre-trained WaveRNN vocoder.

Try it out on Colab:

Version	Colab Link
Forward
Autoregressive

• Installation
• Dataset
• Training
  • Autoregressive
  • Forward
• Prediction
• Model Weights

Make sure you have:

• Python >= 3.6

Install espeak as phonemizer backend (for macOS use brew):

sudo apt-get install espeak

Then install the rest with pip:

pip install -r requirements.txt

Read the individual scripts for more command line arguments.

You can directly use LJSpeech to create the training dataset.

• If training LJSpeech, or if unsure, simply use config/standard
• EDIT PATHS: in data_config.yaml edit the paths to point at your dataset and log folders

Prepare a dataset in the following format:

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

where metadata.csv has the following format: wav_file_name|transcription

python create_dataset.py --config config/standard

python train_autoregressive.py --config config/standard

First use the autoregressive model to create the durations dataset

python extract_durations.py --config config/standard --binary --fix_jumps --fill_mode_next

this will add an additional folder to the dataset folder containing the new datasets for validation and training of the forward model.
If the rhythm of the trained model is off, play around with the flags of this script to fix the durations.

python train_forward.py --config /path/to/config_folder/

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

• To resume training simply use the same configuration files AND --session_name flag, if any
• 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

We log some information that can be visualized with TensorBoard:

tensorboard --logdir /logs/directory/

Predict with either the Forward or Autoregressive model

from utils.config_manager import ConfigManager
from utils.audio import reconstruct_waveform

config_loader = ConfigManager('/path/to/config/', model_kind='forward')
model = config_loader.load_model()
out = model.predict('Please, say something.')

# Convert spectrogram to wav (with griffin lim)
wav = reconstruct_waveform(out['mel'].numpy().T, config=config_loader.config)

• Francesco Cardinale, github: cfrancesco

WaveRNN: we took the data processing from here and use their vocoder to produce the samples.
Erogol and the Mozilla TTS team for the lively exchange on the topic.

See LICENSE for details.