icecream0910/taein-tts
By icecream0910 · ↓ 0 · ♥ 1 · Updated 2024-02-10
What's in this repo
This page lists the files in icecream0910/taein-tts, including hifigan.onnx.
- 1 ONNX models (.onnx files)
- 96 files, including 43 .csv, 9 .def, 8 .exe, 7 .lib, 3 .ipynb, 3 .bin
- Last updated 2024-02-10
ONNX files
| File | Size | Download |
|---|---|---|
| server/models/hifigan-v2/hifigan.onnx | — | Download |
How to use
These ONNX files can be loaded with ONNX Runtime. To inspect a model graph, upload the file to the Netron viewer. Download links point to the original files on Hugging Face.
README
Taein-TTS
Description
Taein-TTS is a project aimed at creating a text-to-speech (TTS) system that reads sentences in my own voice. This repository includes pre-trained models that have been trained using my voice.
Table of Contents
Installation
This README focuses on guiding you through the process of synthesizing speech using pre-trained models, rather than detailing the model training process.
-
Clone the huggingface repository: https://huggingface.co/icecream0910/taein-tts
-
Modify the
run-server.batbatch file in the/serverdirectory to match your actual file paths.For example, if your server folder is at
C:\myown-tts\server, update the file as follows:@echo off setlocal cd /D "%~dp0" set MECAB_KO_DIC_PATH=.\mecab\mecab-ko-dic -r .\mecab\mecabrc set TTS_MODEL_FILE=C:\myown-tts\server\models\glowtts-v2\best_model.pth.tar set TTS_MODEL_CONFIG=C:\myown-tts\server\models\glowtts-v2\config.json set VOCODER_MODEL_FILE=C:\myown-tts\server\models\hifigan-v2\best_model.pth.tar set VOCODER_MODEL_CONFIG=C:\myown-tts\server\models\hifigan-v2\config.json server.exe endlocal -
Update the
glowtts-v2/config.jsonandhifigan-v2/config.jsonfiles in the/server/models/directory with your actual file paths.Ensure you double the backslash (
\\) in the file paths, as shown below:- For
glowtts-v2/config.json:
"stats_path": "C:\\mydata\\tts-server\\models\\glowtts-v2\\scale_stats.npy"- For
hifigan-v2/config.json:
"stats_path": "C:\\mydata\\tts-server\\models\\hifigan-v2\\scale_stats.npy" - For
Usage
To start the TTS server, execute run-server.bat. Once the server is running, you will see the message INFO:werkzeug: * Running on http://0.0.0.0:5000/ (Press CTRL+C to quit) in the command prompt, indicating that the speech synthesis feature is available through the TTS server. To stop the server, press CTRL+C in the command prompt.
API
-
Text preprocessing:
/tts-server/api/process-textSplits sentences and removes special characters to automatically stitch together and playback multi-line sentences as you type.
-
Text Inference:
/tts-server/api/infer-glowttsSynthesizes text to speech. Send the text to be synthesized in the
textparameter of the URL.Example:
http://localhost:5000/tts-server/api/infer-glowtts?text=hello
Text Inference Demo Page
Visit http://localhost:5000/ for a demo.
Contributing
- Fork the repository (https://github.com/icecream0910/myown-tts/fork).
- Create a new branch:
git checkout -b feature/<featureName>. - Commit your changes:
git commit -am 'Add <featureName>'. - Push to the branch:
git push origin feature/<featureName>. - Submit a pull request.
License
This project is licensed under the MIT License.
References
This implementation draws inspiration from the following repositories:
The datasets below are distributed under the CC-BY 2.0 license, with the original text data provided by the Korea Information Society Development Institute's AI Hub, including Korean dialogue text data and Korean-English translation (parallel) corpus text data.
- Korean Corpus for Voice Recording
- SleepingCE Speech Dataset
- Pre-trained Models for SleepingCE Speech Dataset (Glow-TTS)
- Pre-trained Models for SleepingCE Speech Dataset (HiFi-GAN)
- These models were fine-tuned from the model provided by coqui-ai/TTS, trained on the VCTK dataset, available here.
Model info
Created: 2024-02-10
Updated: 2024-02-10
View on Hugging FaceFiles (96)
Related models
More from icecream0910
No other models by this author.