Intergrating Transformer TTS to any project

Pre requirements

• Python >= 3.9
• Clone git repository in your project root directory

git clone https://github.com/okpyjs/TransformerTTS.git

Installation

• Install espeak as phonemizer backend

sudo apt-get install 

• Install virtual environment

virtualenv venv

source venv/bin/activate

• Install dependencies

pip install -r TransformerTTS/requirements.txt

Usage

• Add following code to your script

import sys
sys.path.append('TransformerTTS/')

from data.audio import Audio
from model.factory import tts_ljspeech
model = tts_ljspeech()
audio = Audio.from_config(model.config)
out = model.predict('Please, say something.')

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

Use pre-trained model from command line with

cd TransformerTTS

python predict_tts.py -t "Please, say something."