https://github.com/wenet-e2e/wetts
Production First and Production Ready End-to-End Text-to-Speech Toolkit
https://github.com/wenet-e2e/wetts
Last synced: about 1 year ago
JSON representation
Production First and Production Ready End-to-End Text-to-Speech Toolkit
- Host: GitHub
- URL: https://github.com/wenet-e2e/wetts
- Owner: wenet-e2e
- License: apache-2.0
- Created: 2022-02-19T02:09:40.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2024-05-30T05:34:34.000Z (about 2 years ago)
- Last Synced: 2025-03-29T06:51:34.601Z (about 1 year ago)
- Language: Python
- Size: 3.43 MB
- Stars: 384
- Watchers: 15
- Forks: 59
- Open Issues: 24
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# WeTTS
Production First and Production Ready End-to-End Text-to-Speech Toolkit
## Install
### Install python package
``` sh
pip install git+https://github.com/wenet-e2e/wetts.git
```
**Command-line usage** (use `-h` for parameters):
``` sh
wetts --text "今天天气怎么样" --wav output.wav
```
**Python programming usage**:
``` python
import wetts
# TODO
```
### Install for development & deployment
We suggest to install WeTTS with Anaconda or Miniconda.
Clone this repo:
```sh
git clone https://github.com/wenet-e2e/wetts.git
```
Create the environment:
```bash
conda create -n wetts python=3.8 -y
conda activate wetts
pip install -r requirements.txt
```
## Roadmap
We mainly focus on end to end, production, and on-device TTS. We are going to use:
* backend: end to end model, such as:
* [VITS](https://arxiv.org/pdf/2106.06103.pdf)
* frontend:
* Text Normalization: [WeTextProcessing](https://github.com/wenet-e2e/WeTextProcessing)
* Prosody & Polyphones: [Unified Mandarin TTS Front-end Based on Distilled BERT Model](https://arxiv.org/pdf/2012.15404.pdf)
## Dataset
We plan to support a variaty of open source TTS datasets, include but not limited to:
* [Baker](https://www.data-baker.com/data/index/TNtts), Chinese Standard Mandarin Speech corpus open sourced by Data Baker.
* [AISHELL-3](https://openslr.org/93), a large-scale and high-fidelity multi-speaker Mandarin speech corpus.
* [Opencpop](https://wenet.org.cn/opencpop), Mandarin singing voice synthesis (SVS) corpus open sourced by Netease Fuxi.
## Pretrained Models
| Dataset | Language | Checkpoint Model | Runtime Model |
| -------------- | -------- | ---------------- | ------------- |
| Baker | CN | [BERT](https://wenet.org.cn/downloads?models=wetts&version=baker_bert_exp.tar.gz) | [BERT](https://wenet.org.cn/downloads?models=wetts&version=baker_bert_onnx.tar.gz) |
| Multilingual | CN | [VITS](https://wenet.org.cn/downloads?models=wetts&version=multilingual_vits_v3_exp.tar.gz) | [VITS](https://wenet.org.cn/downloads?models=wetts&version=multilingual_vits_v3_onnx.tar.gz) |
## Runtime
We plan to support a variaty of hardwares and platforms, including:
* x86
* Android
* Raspberry Pi
* Other on-device platforms
``` bash
export GLOG_logtostderr=1
export GLOG_v=2
cd runtime/onnxruntime
cmake -B build -DCMAKE_BUILD_TYPE=Release
cmake --build build
./build/bin/tts_main \
--frontend_flags baker_bert_onnx/frontend.flags \
--vits_flags multilingual_vits_v3_onnx/vits.flags \
--sname baker \
--text "hello我是小明。" \
--wav_path audio.wav
```
## Discussion & Communication
For Chinese users, you can aslo scan the QR code on the left to follow our offical account of WeNet.
We created a WeChat group for better discussion and quicker response.
Please scan the personal QR code on the right, and the guy is responsible for inviting you to the chat group.
| 
| 
|
| ---- | ---- |
Or you can directly discuss on [Github Issues](https://github.com/wenet-e2e/wetts/issues).
## Acknowledgement
1. We borrow a lot of code from [vits](https://github.com/jaywalnut310/vits) for VITS implementation.
2. We refer [PaddleSpeech](https://github.com/PaddlePaddle/PaddleSpeech) for `pinyin` lexicon generation.