Shona TTS

🚀 Shona Text-to-Speech

This repository offers a text-to-speech (TTS) model checkpoint for the Shona (sna) language, facilitating the conversion of text into natural-sounding speech.

🚀 Quick Start

To start using the Shona Text-to-Speech model, first install the necessary libraries:

pip install --upgrade transformers accelerate

Then, you can run inference with the following Python code:

# Load model directly
from transformers import AutoTokenizer, AutoModelForTextToWaveform

tokenizer = AutoTokenizer.from_pretrained("Fastino06/ff")
model = AutoModelForTextToWaveform.from_pretrained("Fastino06/ff")

text = "some example text in the Shona language"
inputs = tokenizer(text, return_tensors="pt")

with torch.no_grad():
    output = model(**inputs).waveform

The resulting waveform can be saved as a .wav file:

import scipy

scipy.io.wavfile.write("fassy.wav", rate=model.config.sampling_rate, data=output)

Or displayed in a Jupyter Notebook / Google Colab:

from IPython.display import Audio

Audio(output, rate=model.config.sampling_rate)

✨ Features

• Language Support: Specifically designed for the Shona language, enabling high - quality text - to - speech conversion.
• Model Architecture: Based on the SpeechT5 model, fine - tuned for optimal performance in Shona TTS.

📦 Installation

pip install --upgrade transformers accelerate

💻 Usage Examples

Basic Usage

# Load model directly
from transformers import AutoTokenizer, AutoModelForTextToWaveform

tokenizer = AutoTokenizer.from_pretrained("Fastino06/ff")
model = AutoModelForTextToWaveform.from_pretrained("Fastino06/ff")

text = "some example text in the Shona language"
inputs = tokenizer(text, return_tensors="pt")

with torch.no_grad():
    output = model(**inputs).waveform

Saving the Output

import scipy

scipy.io.wavfile.write("fassy.wav", rate=model.config.sampling_rate, data=output)

Displaying in Notebook

from IPython.display import Audio

Audio(output, rate=model.config.sampling_rate)

📚 Documentation

Model Details

📄 License

This model is licensed under the CC - BY - NC - 4.0 license.

BibTex citation

This model was developed by Fastino Mateteva