|
|
--- |
|
|
license: apache-2.0 |
|
|
language: |
|
|
- hi |
|
|
size_categories: |
|
|
- 10K<n<100K |
|
|
--- |
|
|
# 🗣️ Hindi Text-to-Speech (TTS) Model |
|
|
|
|
|
This repository hosts a Hindi TTS model trained on a custom Hindi dataset. The model converts Hindi text into natural-sounding speech and is suitable for various applications such as voice assistants, audiobooks, and accessibility tools. |
|
|
|
|
|
## 📌 Model Details |
|
|
|
|
|
- **Model Type**: [Insert TTS architecture, e.g., Tacotron2 + HiFi-GAN / VITS] |
|
|
- **Language**: Hindi (`hi`) |
|
|
- **Dataset**: Custom Hindi speech dataset (uploaded by the author) |
|
|
- **Sampling Rate**: 22050 Hz (or specify your sample rate) |
|
|
- **License**: [Apache License 2.0 x, e.g., MIT, CC-BY-4.0, etc.] |
|
|
|
|
|
## 🏁 Usage |
|
|
|
|
|
### Inference using `transformers` + `datasets` (if applicable) |
|
|
|
|
|
```python |
|
|
from transformers import pipeline |
|
|
|
|
|
# Load the model |
|
|
tts = pipeline("text-to-speech", model="Saurabh4509/TSS_Hindi_Data") |
|
|
|
|
|
# Generate speech |
|
|
output = tts("नमस्ते, आप कैसे हैं?") |
|
|
|
|
|
# Save to file |
|
|
with open("output.wav", "wb") as f: |
|
|
f.write(output["audio"]) |
