Text-to-Speech (TTS) with PyTorch

About • Installation • How To Use • Credits • License

About

This repository contains an implemetation of HiFi-GAN with PyTorch.

Installation

Follow these steps to install the project:

0. (Optional) Create and activate new environment using conda or venv (+pyenv).

   a. conda version:

   # create env
   conda create -n project_env python=PYTHON_VERSION
   
   # activate env
   conda activate project_env

   b. venv (+pyenv) version:

   # create env
   ~/.pyenv/versions/PYTHON_VERSION/bin/python3 -m venv project_env
   
   # alternatively, using default python version
   python3 -m venv project_env
   
   # activate env
   source project_env

1. Install all required packages

   pip install -r requirements.txt

2. Install pre-commit:

   pre-commit install

How To Use

To train a model, run the following command:

python3 train.py -cn=hifigan

How To Download

To download pretrained model, use:

gdown https://drive.google.com/uc?id=1n9DVZznWy49nKiSljAbqdQvNiZ_VcPOa

How To Evaluate

To synthesize an audio from audio, your dataset should follow this structure:

NameOfTheDirectoryWithUtterances
└── transcriptions
    ├── UtteranceID1.wav
    ├── UtteranceID2.wav
    .
    .
    .
    └── UtteranceIDn.wav

To get predictions, run

python3 synthesize.py -cn=from_audio '+datasets.test.audio_dir=<PATH-TO-DIR>' 'inferencer.from_pretrained=<PATH-TO-PRETRAINED-MODEL>'

To synthesize an audio from text, your dataset should follow this structure:

NameOfTheDirectoryWithUtterances
└── transcriptions
    ├── UtteranceID1.txt
    ├── UtteranceID2.txt
    .
    .
    .
    └── UtteranceIDn.txt

To get predictions, run

python3 synthesize.py -cn=from_text '+datasets.test.data_dir=<PATH-TO-DIR>' 'inferencer.from_pretrained=<PATH-TO-PRETRAINED-MODEL>'

If you want to pass text from cli, run:

python3 synthesize.py -cn=from_cli '+datasets.test.index=[{text: "<YOUR-TEXT>", path: "text.txt", audio_len: 0}]' 'inferencer.from_pretrained=<PATH-TO-PRETRAINED-MODEL>'

Report

Wandb report is available here.

Credits

This repository is based on a PyTorch Project Template.

License