B station open source project IndexTTS deployment tutorial

jsjs · Posted 6 days ago

First of all, Windows is not supported, Windows installation dependency will report an error: https: //github.com/index-tts/index-tts/issues/28

Recap:

"Kokoro-82M", a text-to-speech TTS model that has recently exploded in popularity.
h ttps:// www.itsvse.com/thread-10928-1-1.html

IndexTTS

Industrial-grade Controlled and Efficient Zero-Sample Text-to-Speech System

IndexTTS is a GPT-style text-to-speech (TTS) model based on XTTS and Tortoise algorithms. It is capable of correcting the pronunciation of Chinese characters through pinyin and controlling pauses at arbitrary positions through punctuation. We have enhanced several modules of the system, including improved speaker-conditional feature representation and integration of BigVGAN2 to optimize audio quality. Our system is trained on tens of thousands of hours of data to achieve optimal performance, outperforming popular TTS systems such as XTTS, CosyVoice2, Fish-Speech, and F5-TTS.

Project address: https: //github.com/index-tts/index-tts
Online experience: https: //huggingface.co/spaces/IndexTeam/IndexTTS
Model download: https: //huggingface.co/IndexTeam/IndexTTS-1.5

Model effect: IndexTTS-1.5 > IndexTTS

The overview of IndexTTS is as follows:

IndexTTS Deployment

Hardware: CPU N150, RAM: 16G, System: Rocky Linux 9

To install conda, please refer to the previous article.

Creating an environment with conda

Log in to be visible.

Download the model

Login to see.

Install dependencies and clone index-tts project

Log in to be visible.

Reasoning Test

Log in to be visible.

You can see that the output.wav voice file has been generated. (It took a long time because my CPU performance is weak)

(finished)

		Auto Login	Retrieve password
Retrieve Password			Register

B station open source project IndexTTS deployment tutorial

Related posts