https://github.com/seanghay/klea
An open-source Khmer Word to Speech Model. Just single word not sentence!
https://github.com/seanghay/klea
cambodia khmer khmer-synthesis tts vits
Last synced: 3 months ago
JSON representation
An open-source Khmer Word to Speech Model. Just single word not sentence!
- Host: GitHub
- URL: https://github.com/seanghay/klea
- Owner: seanghay
- License: mit
- Created: 2023-10-05T13:02:21.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2024-02-03T14:09:58.000Z (over 1 year ago)
- Last Synced: 2024-02-04T14:54:25.962Z (over 1 year ago)
- Topics: cambodia, khmer, khmer-synthesis, tts, vits
- Language: Python
- Homepage: https://huggingface.co/spaces/seanghay/KLEA
- Size: 610 KB
- Stars: 2
- Watchers: 1
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# KLEA
An open-source Khmer Word to Speech Model. Just single word not sentence!
### 1. Setup
```shell
pip install -r requirements.txt
```
### 2. Download Checkpoint
[G_60000.pth](https://huggingface.co/spaces/seanghay/KLEA/resolve/main/G_60000.pth)
```shell
wget https://huggingface.co/spaces/seanghay/KLEA/resolve/main/G_60000.pth
```
Place the checkpoint in the current directory.
### 3. Inference
```shell
python infer.py "មនុស្សខ្មែរ"
```
This will output a file called `audio.wav` in the current directory. Output audio sample rate is 22.05 kHz.
### Gradio
```
python app.py
```
### Colab
### Dataset
This model was trained on kheng.info dataset. You can find it on http://kheng.info or at https://hf.co/datasets/seanghay/khmer_kheng_info_speech
## Reference
- [VITS: Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech](https://github.com/jaywalnut310/vits)
- [kheng.info](https://kheng.info/about/) is an online audio dictionary for the Khmer language with over 3000 recordings. Kheng.info is backed by multiple dictionaries and a large text corpus, and supports search in English and Khmer with search results ordered by word frequency.