MMS TTS THAI FEMALEV1

🚀 Thai Text-to-Speech Model

This project provides a Thai text-to-speech model, which is finetuned using audio recordings from Play.ht and offers various usage methods.

🚀 Quick Start

This model uses audio recordings from Play.ht: Play.ht to finetune the model.

📦 Installation

• Finetune the model: You can find the GitHub code for finetuning the model here: Finetune GitHub Code.
• Train the model on Google Colab: Use this Finetune-colab to train the speech model on Google Colab by yourself.
• Run on a local computer: You can run the model on your local computer following the guide at Local Usage.

💻 Usage Examples

Basic Usage

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

tokenizer = VitsTokenizer.from_pretrained("VIZINTZOR/VIZINTZOR/MMS-TTS-THAI-FEMALEV1",cache_dir="./mms")
model = VitsModel.from_pretrained("VIZINTZOR/VIZINTZOR/MMS-TTS-THAI-FEMALEV1",cache_dir="./mms")

inputs = tokenizer(text="สวัสดีค่ะ นี่คือเสียงพูดภาษาไทย", 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)