Step-by-step tutorial on how to clone your voice using python and ai

Mark Caggiano
4 min readMay 9, 2023

Cloning a voice using Python and AI involves training a model on your voice recordings and generating new speech based on the learned patterns. Here’s a step-by-step tutorial on how to achieve this using Tacotron 2 and WaveGlow, popular models for text-to-speech synthesis:

Step 1: Set up the Environment

  1. Install Python: Make sure you have Python 3.x installed on your system.
  2. Install the required packages: Open a terminal or command prompt and run the following commands to install the necessary packages:
pip install numpy torch==1.4.0 librosa unidecode inflect scipy
pip install git+
pip install unidecode
pip install pillow
pip install tensorboardX

Install Tacotron 2 and WaveGlow: Clone the Tacotron 2 and WaveGlow repositories from GitHub using the following commands:

git clone
git clone

Install additional requirements: Navigate to the cloned Tacotron 2 repository and install the additional requirements by running:

cd tacotron2
pip install -r requirements.txt