Persian-tts-coqui

Persian/Farsi text to speech(TTS) training using coqui tts (Online demo : )
This repository contains sample codes for training text to speech models
Feel free to ask your questions issues

این مخزن شامل نمونه کدهای لازم برای آموزش مدل های متن به صوت فارسی است سوالاتتان را در issues مطرح کنید

How to train ?

Sample codes and notebooks are available at recepies folder

نمونه کد ها و نوت بوک ها در پوشه recepies موجود هستند

Pretrained models

These are models you can use to test or finetune

مدل هایی که می توانید برای امتحان کردن یا finetune کردن از آنها استفاده کنید

Model	Dataset
vits female (best)	persian-tts-dataset-famale
vits male1 (best)	persian-tts-dataset-male
vits female1	ParsiGoo
vits male	persian-tts-dataset
glowtts female	persian-tts-dataset-famale
glowtts male	persian-tts-dataset
tacotron2 female	persian-tts-dataset-famale
Hifigan	persian-tts-dataset-famale
Wavernn	persian-tts-dataset-famale

• Share your trained models here

• 🤗 huggingface Demo https://huggingface.co/spaces/Kamtera/Persian-tts-CoquiTTS

Datasets

Models trained on these datasets :

• https://www.kaggle.com/datasets/magnoliasis/persian-tts-dataset

• https://www.kaggle.com/datasets/magnoliasis/persian-tts-dataset-famale

• https://www.kaggle.com/datasets/magnoliasis/persian-tts-dataset-male

• https://github.com/karim23657/ParsiGoo

• If you'v created a dataset or found any good datasets on the web you can share with us here.

Test trained model

• install TTS and Espeak

pip install -q TTS
sudo apt-get -y install espeak-ng

• predict one text from commandline

tts --text "شیش سیخ جیگر" --model_path "best_model.ckpt" --config_path "config.json"

• From python API

from TTS.api import TTS
tts=TTS(model_path="https://huggingface.co/Kamtera/persian-tts-male1-vits/resolve/main/checkpoint_88000.pth",
        config_path="https://huggingface.co/Kamtera/persian-tts-male1-vits/resolve/main/config.json")
tts.tts_to_file(".زندگی فقط یک بار است؛ از آن به خوبی استفاده کن",file_path='output.wav')

• Another way :

from TTS.config import load_config
from TTS.utils.manage import ModelManager
from TTS.utils.synthesizer import Synthesizer


model_path ="config.json"  # Absolute path to the model checkpoint.pth
config_path ="best_model.pth" # Absolute path to the model config.json

text=".زندگی فقط یک بار است؛ از آن به خوبی استفاده کن"

synthesizer = Synthesizer(
    model_path, config_path
)
wavs = synthesizer.tts(text)
synthesizer.save_wav(wavs, 'sp.wav')

usefull links

If you want more persian text to speech tools: ابزار ها صوت به متن فارسی بیشتر :

• https://github.com/karim23657/awesome-Persian-Speech#tts-tools

Any issue about tts library:

• https://github.com/coqui-ai/TTS

❓ Questions

🤩 If you have any questions, feedback, or suggestions, feel free to reach out to me at alias@karim23657.anonaddy.com ! I'm always happy to help.

🤩 در صورتی که سوال، بازخورد یا پیشنهادی دارید، با من از طریق ایمیل alias@karim23657.anonaddy.com در تماس باشید! خوشحال میشم که بتونم کمک کنم.