https://github.com/wannaphong/ttsmms

TTS with The Massively Multilingual Speech (MMS) project
https://github.com/wannaphong/ttsmms

library python text-to-speech

Last synced: 4 days ago
JSON representation

TTS with The Massively Multilingual Speech (MMS) project

• Host: GitHub
• URL: https://github.com/wannaphong/ttsmms
• Owner: wannaphong
• License: mit
• Created: 2023-05-23T05:14:12.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2024-07-12T13:22:54.000Z (6 months ago)
• Last Synced: 2024-12-15T00:03:55.177Z (11 days ago)
• Topics: library, python, text-to-speech
• Language: Python
• Homepage:
• Size: 3.28 MB
• Stars: 226
• Watchers: 7
• Forks: 37
• Open Issues: 12
• Metadata Files:
  • Readme: README.md
  • License: LICENSE
  • Support: support_list.txt

Awesome Lists containing this project

README

        
# ttsmms

Text-to-speech with The Massively Multilingual Speech (MMS) project

This project want to help you for use Text-to-speech model from MMS project in Python.

Support 1,107 Languages! (See support_list.txt)

- VITS: [GitHub](https://github.com/jaywalnut310/vits)

- MMS: Scaling Speech Technology to 1000+ languages: [GitHub](https://github.com/facebookresearch/fairseq/tree/main/examples/mms)

[Google colab](https://colab.research.google.com/github/wannaphong/ttsmms/blob/main/notebook/test.ipynb)

**Don't forget the TTS model in MMS project use CC-BY-NC license.**

## Install

> pip install ttsmms

## Usage

First, you need to download the model by

```python

from ttsmms import download

dir_path = download("eng","./data") # lang_code, dir for save model

```

or download file by yourself

**Linux/Mac**

1. download

> curl https://dl.fbaipublicfiles.com/mms/tts/lang_code.tar.gz --output lang_code.tar.gz

2. extract a tar ball archive.

**Linux/Mac**

> mkdir -p data && tar -xzf lang_code.tar.gz -C data/

and use code in python :D

```python

from ttsmms import TTS

tts=TTS(dir_path) # or "model_dir_path" your path dir that extract a tar ball archive

wav=tts.synthesis("txt")

# output:

# {

#    "x":array(wav array),

#    "sampling_rate": 16000

# }

tts.synthesis("txt",wav_path="example.wav")

# output: example.wav file

```