initial app file development

README.md

@@ -1,5 +1,5 @@
 ---
-title: Chatterbox
+title: Chatterbox Multilingual TTS Demo
 emoji: 👁
 colorFrom: purple
 colorTo: pink
@@ -9,4 +9,44 @@ app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Chatterbox Multilingual TTS Demo
+
+This demo showcases the multilingual capabilities of the Chatterbox TTS model, supporting both English and Chinese languages.
+
+## Features
+
+- Text-to-speech generation for English and Chinese
+- Gradio web interface for easy interaction
+- Real-time audio generation and playback
+- Example texts for quick testing
+
+## Requirements
+
+- Python 3.8 or higher
+- Required Python packages (automatically installed by Hugging Face):
+  - chatterbox-tts
+  - gradio
+  - torchaudio
+  - torch
+
+## Usage
+
+1. Enter text in the input box
+2. Select the language (English or Chinese)
+3. Click "Generate Speech"
+4. Listen to the generated audio
+
+## Supported Languages
+
+- English
+- Chinese
+
+## Examples
+
+The interface includes example texts for both languages to help you get started quickly.
+
+## Notes
+
+- The first generation may take a moment as the model loads
+- Subsequent generations will be faster
+- For best results, use clear and properly punctuated text

app.py

@@ -1,7 +1,80 @@
 import gradio as gr
+import torchaudio as ta
+import torch
+import tempfile
+import os
+from chatterbox.mtl_tts import ChatterboxMultilingualTTS
 
-def greet(name):
-    return "Hello " + name + "!!"
+# Initialize the multilingual model
+model = ChatterboxMultilingualTTS.from_pretrained(device="cuda" if torch.cuda.is_available() else "cpu")
 
-demo = gr.Interface(fn=greet, inputs="text", outputs="text")
-demo.launch()
+def generate_speech(text, language):
+    """
+    Generate speech from text using Chatterbox multilingual TTS
+    
+    Args:
+        text (str): Text to convert to speech
+        language (str): Language code ('en' for English, 'zh' for Chinese)
+    
+    Returns:
+        str: Path to the generated audio file
+    """
+    # Map language codes to full names for Chatterbox
+    language_map = {
+        "English": "en",
+        "Chinese": "zh"
+    }
+    
+    language_id = language_map.get(language, "en")
+    
+    # Generate speech using Chatterbox
+    wav = model.generate(text, language_id=language_id)
+    
+    # Save to a temporary file
+    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as tmp_file:
+        ta.save(tmp_file.name, wav, model.sr)
+        return tmp_file.name
+
+# Create Gradio interface
+with gr.Blocks(title="Chatterbox Multilingual TTS Demo") as demo:
+    gr.Markdown("# Chatterbox Multilingual TTS Demo")
+    gr.Markdown("This demo uses Chatterbox to generate speech in English and Chinese.")
+    
+    with gr.Row():
+        with gr.Column():
+            text_input = gr.Textbox(
+                label="Input Text",
+                placeholder="Enter text to convert to speech...",
+                lines=3
+            )
+            language_selection = gr.Radio(
+                choices=["English", "Chinese"],
+                value="English",
+                label="Language"
+            )
+            generate_btn = gr.Button("Generate Speech")
+        
+        with gr.Column():
+            audio_output = gr.Audio(label="Generated Speech", type="filepath")
+    
+    # Examples
+    gr.Examples(
+        examples=[
+            ["Hello, welcome to the Chatterbox multilingual demo. This is an English example.", "English"],
+            ["你好，欢迎来到Chatterbox多语言演示。这是一个中文示例。", "Chinese"]
+        ],
+        inputs=[text_input, language_selection],
+        outputs=audio_output,
+        fn=generate_speech,
+        cache_examples=True
+    )
+    
+    # Connect the generate button to the function
+    generate_btn.click(
+        fn=generate_speech,
+        inputs=[text_input, language_selection],
+        outputs=audio_output
+    )
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,4 @@
+chatterbox-tts
+gradio>=5.44.1
+torchaudio
+torch