MMS-TTS-THAI-MALE-NARRATOR Open-source Thai Text-to-Speech Model - Free Deployment, Great Male Narrator Effect

Home

MMS TTS THAI MALE NARRATOR

Developed by VIZINTZOR

This is a Thai text-to-speech model based on the VITS architecture, specifically fine-tuned for male narration style.

Speech Synthesis

Safetensors

Other#Thai speech synthesis #Male narrator voice #Localized fine-tuning

Downloads 205

Release Time : 1/5/2025

Model Overview

The model can convert Thai text into natural and fluent male narration speech, suitable for speech synthesis, audio content creation, and other scenarios.

Model Features

Thai speech synthesis

Speech synthesis capability optimized specifically for Thai, providing natural male narration voice.

Based on VITS architecture

Utilizes the advanced VITS end-to-end text-to-speech architecture for high-quality speech synthesis.

Local deployment support

Provides guidelines for local computer usage and fine-tuning code, supporting personalized customization.

Model Capabilities

Thai text-to-speech

Speech synthesis

Voice style conversion (male narration)

Use Cases

Content creation

Audiobook production

Convert Thai text content into natural speech for audiobook production.

High-quality male narration voice output.

Educational content dubbing

Provide Thai voice-over for educational videos or courseware.

Clear and professional instructional voice.

Assistive technology

Visual impairment assistance

Convert Thai text content into speech to help visually impaired individuals access information.

Understandable speech output.

🚀 Thai Text-to-Speech Model

This project provides a Thai text-to-speech model. It uses audio recordings from Play.ht for fine-tuning the model and offers various ways to train and use the model.

🚀 Quick Start

Prerequisites

Make sure you have Python and the necessary libraries installed.

Installation

You can use the model on your local computer by referring to the repository: VachanaTTS

Usage

import torch
from transformers import VitsTokenizer, VitsModel, set_seed
import scipy

tokenizer = VitsTokenizer.from_pretrained("VIZINTZOR/MMS-TTS-THAI-MALE-NARRATOR",cache_dir="./mms")
model = VitsModel.from_pretrained("VIZINTZOR/MMS-TTS-THAI-MALE-NARRATOR",cache_dir="./mms")

inputs = tokenizer(text="สวัสดีครับ นี่คือการบรรยายภาษาไทย", return_tensors="pt")

set_seed(456)  # make deterministic

with torch.no_grad():
   outputs = model(**inputs)

waveform = outputs.waveform[0]

# Convert PyTorch tensor to NumPy array
waveform_array = waveform.numpy()

scipy.io.wavfile.write("techno_output.wav", rate=model.config.sampling_rate, data=waveform_array)

✨ Features

Fine-tuning: Fine-tune the model using audio recordings from Play.ht. You can find the fine-tuning code on GitHub.
Training on Google Colab: Train the speech model on your own using Finetune-colab.
Local Usage: Use the model on your local computer with the help of VachanaTTS.

📦 Installation

To use the model on your local computer, refer to the repository: VachanaTTS

💻 Usage Examples

Basic Usage

import torch
from transformers import VitsTokenizer, VitsModel, set_seed
import scipy

tokenizer = VitsTokenizer.from_pretrained("VIZINTZOR/MMS-TTS-THAI-MALE-NARRATOR",cache_dir="./mms")
model = VitsModel.from_pretrained("VIZINTZOR/MMS-TTS-THAI-MALE-NARRATOR",cache_dir="./mms")

inputs = tokenizer(text="สวัสดีครับ นี่คือการบรรยายภาษาไทย", return_tensors="pt")

set_seed(456)  # make deterministic

with torch.no_grad():
   outputs = model(**inputs)

waveform = outputs.waveform[0]

# Convert PyTorch tensor to NumPy array
waveform_array = waveform.numpy()

scipy.io.wavfile.write("techno_output.wav", rate=model.config.sampling_rate, data=waveform_array)

Property	Details
Model Type	Thai text-to-speech model
Training Data	Audio recordings from Play.ht: https://play.ht/

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご