app.py

"""
ChatterBox Bengali TTS Space
Fine-tuned Bengali TTS model for text-to-speech synthesis
"""

import sys
import os
import tempfile
import gradio as gr
import torch
from huggingface_hub import snapshot_download
import torchaudio

# Add chatterbox-finetuning to path
chatterbox_finetuning_path = os.path.join(os.path.dirname(__file__), "chatterbox-finetuning")
sys.path.insert(0, chatterbox_finetuning_path)

from src.chatterbox_.tts import ChatterboxTTS

# Model configuration
MODEL_ID = "BosonLab/chatterbox-bangla"
DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

# Load model and tokenizer
print(f"Loading model from {MODEL_ID}...")
print(f"Using device: {DEVICE}")

try:
    # Download model from Hugging Face
    model_dir = snapshot_download(MODEL_ID)
    print(f"Model downloaded to: {model_dir}")

    # Load ChatterBox TTS model
    model = ChatterboxTTS.from_local(model_dir, device=DEVICE)
    print("Model loaded successfully!")

except Exception as e:
    print(f"Error loading model: {e}")
    raise


def generate_speech(text, reference_audio=None):
    """
    Generate speech from text using the fine-tuned model

    Args:
        text: Bengali text to convert to speech
        reference_audio: Path to reference audio file for voice cloning

    Returns:
        Tuple of (audio, metadata)
    """
    try:
        # Create temporary directory for output
        temp_dir = tempfile.mkdtemp()

        # Generate speech with voice cloning if reference audio is provided
        if reference_audio is not None:
            wav = model.generate(text, audio_prompt_path=reference_audio)
            output_path = os.path.join(temp_dir, "output.wav")
        else:
            wav = model.generate(text)
            output_path = os.path.join(temp_dir, "output.wav")

        return (model.sr, wav.squeeze(0).numpy()),  f"Generated {len(text)} characters of Bengali text"

        # Save audio to file
        # torchaudio.save(output_path, wav, model.sr)

        # return output_path, f"Generated {len(text)} characters of Bengali text"

    except Exception as e:
        print(f"Error generating speech: {e}")
        return None, f"Error: {str(e)}"


# Create Gradio interface
with gr.Blocks(title="ChatterBox Bengali TTS", theme=gr.themes.Soft()) as demo:

    with gr.Row():
        with gr.Column(scale=1):
            text_input = gr.Textbox(
                label="Bengali Text",
                placeholder="Enter Bengali text here...",
                lines=5,
                max_lines=10
            )

            reference_audio = gr.Audio(
                label="Reference Audio (Optional for Voice Cloning)",
                type="filepath"
            )

            generate_btn = gr.Button("Generate Speech", variant="primary")

        with gr.Column(scale=1):
            audio_output = gr.Audio(label="Generated Speech")
            metadata_output = gr.Textbox(label="Metadata", lines=3)

    # Event handlers
    generate_btn.click(
        fn=generate_speech,
        inputs=[text_input, reference_audio],
        outputs=[audio_output, metadata_output]
    )


if __name__ == "__main__":
    demo.launch(
        server_name="0.0.0.0",
        server_port=7860,
        share=False
    )