README.md

---
datasets:
- mesolitica/TTS
language:
- ms
---
# MeloTTS MS

MeloTTS continue train on MS, forked at https://github.com/mesolitica/MeloTTS-MS

**still on training**.

## Checkpoints

We uploaded full checkpoints with optimizer states at [checkpoints](checkpoints).

## Dataset

We train on [Mesolitica/TTS](https://huggingface.co/datasets/mesolitica/TTS).

## requirements

```bash
pip3 install git+https://github.com/mesolitica/MeloTTS-MS
pip3 install git+https://github.com/mesolitica/malaya-speech malaya phonemizer
apt install espeak -y
```

## how to

```python
from huggingface_hub import snapshot_download
from from melo.api import TTS

ckpt_path = hf_hub_download(repo_id='mesolitica/MeloTTS-MS', filename='model.pth')
config_path = hf_hub_download(repo_id='mesolitica/MeloTTS-MS', filename='config.json')
model = TTS(language='MS', config_path=config_path, ckpt_path=ckpt_path)

speaker_id = {
  'Husein': 0,
  'Shafiqah Idayu': 1,
  'Anwar Ibrahim': 2,
}

r = model.tts_to_file('Husein sangatlah comel dan wangi.', 0, split = True, sdp_ratio = 0, noise_scale = 0.667)
```