Update app.py

app.py

@@ -1,50 +1,111 @@
 import gradio as gr
 from supertonic import TTS
 import os
+import numpy as np
 
-# Initialize TTS
-tts = TTS(auto_download=True)
+# Initialize TTS - auto_download=True handles the HF model fetching automatically
+# This runs on CPU by default via ONNX
+try:
+    tts = TTS(auto_download=True)
+except Exception as e:
+    print(f"Error initializing TTS: {e}")
 
+# All available voices found in the repository tree
 VOICES = ["M1", "M2", "M3", "M4", "M5", "F1", "F2", "F3", "F4", "F5"]
+
+# The 31 supported languages
 LANGUAGES = {
-    "English": "en", "Korean": "ko", "Japanese": "ja", "Spanish": "es",
-    "French": "fr", "German": "de", "Hindi": "hi", "Italian": "it"
+    "English": "en", "Korean": "ko", "Japanese": "ja", "Arabic": "ar",
+    "Bulgarian": "bg", "Czech": "cs", "Danish": "da", "German": "de",
+    "Greek": "el", "Spanish": "es", "Estonian": "et", "Finnish": "fi",
+    "French": "fr", "Hindi": "hi", "Croatian": "hr", "Hungarian": "hu",
+    "Indonesian": "id", "Italian": "it", "Lithuanian": "lt", "Latvian": "lv",
+    "Dutch": "nl", "Polish": "pl", "Portuguese": "pt", "Romanian": "ro",
+    "Russian": "ru", "Slovak": "sk", "Slovenian": "sl", "Swedish": "sv",
+    "Turkish": "tr", "Ukrainian": "uk", "Vietnamese": "vi"
 }
 
 def generate_speech(text, voice, language_name):
-    lang_code = LANGUAGES[language_name]
-    style = tts.get_voice_style(voice_name=voice)
-    wav, duration = tts.synthesize(text, voice_style=style, lang=lang_code)
+    if not text.strip():
+        raise gr.Error("Please enter some text.")
+        
+    try:
+        lang_code = LANGUAGES[language_name]
+        
+        # Get the voice style object
+        style = tts.get_voice_style(voice_name=voice)
+        
+        # Synthesize (Returns wav data and a numpy array for duration)
+        wav, duration = tts.synthesize(text, voice_style=style, lang=lang_code)
+        
+        # Save to a temporary path
+        output_path = "output.wav"
+        tts.save_audio(wav, output_path)
+        
+        # FIX: Convert numpy.ndarray duration to float for f-string compatibility
+        readable_duration = float(duration)
+        
+        return output_path, f"Generation Successful! \nDuration: {readable_duration:.2f}s"
     
-    output_path = "output.wav"
-    tts.save_audio(wav, output_path)
-    return output_path, f"Duration: {duration:.2f}s"
+    except Exception as e:
+        raise gr.Error(f"Generation failed: {str(e)}")
 
-# Gradio 6 focuses on cleaner layout and updated theme handling
+# Define the Gradio Interface
 with gr.Blocks(theme='soft', title="Supertonic 3 TTS") as demo:
-    gr.Markdown("## 🎙️ Supertonic 3 TTS (Secure & Optimized)")
+    gr.Markdown("# 🎙️ Supertonic 3: Multilingual TTS")
+    gr.Markdown("An on-device, lightweight Text-to-Speech system by **Supertone**. Running on CPU via ONNX.")
     
     with gr.Row():
         with gr.Column(scale=1):
             input_text = gr.Textbox(
                 label="Input Text", 
-                placeholder="Type here...",
-                value="Hello! I am now running on a secure version of Gradio.",
-                lines=3
+                placeholder="Type your message here...",
+                value="The train delay was announced at 4:45 PM <breath> due to track maintenance.",
+                lines=4
             )
+            
             with gr.Row():
-                voice_opt = gr.Dropdown(choices=VOICES, value="M1", label="Voice")
-                lang_opt = gr.Dropdown(choices=list(LANGUAGES.keys()), value="English", label="Language")
-            btn = gr.Button("Generate", variant="primary")
+                voice_opt = gr.Dropdown(
+                    choices=VOICES, 
+                    value="M1", 
+                    label="Voice Style"
+                )
+                lang_opt = gr.Dropdown(
+                    choices=sorted(list(LANGUAGES.keys())), 
+                    value="English", 
+                    label="Language"
+                )
+            
+            btn = gr.Button("Synthesize Speech", variant="primary")
             
         with gr.Column(scale=1):
-            audio_output = gr.Audio(label="Result", type="filepath")
-            stats = gr.Textbox(label="Metadata", interactive=False)
+            audio_output = gr.Audio(label="Synthesized Audio", type="filepath")
+            status_box = gr.Textbox(label="Status", interactive=False)
+            
+            gr.HTML("""
+                <div style="margin-top: 20px; text-align: center;">
+                    <p>Supported Expression Tags: <code>&lt;laugh&gt;</code>, <code>&lt;breath&gt;</code>, <code>&lt;sigh&gt;</code></p>
+                    <a href="https://huggingface.co/Supertone/supertonic-3">
+                        <img src="https://img.shields.io/badge/Model-Supertone%20Supertonic--3-blue?logo=huggingface" alt="Model Page">
+                    </a>
+                </div>
+            """)
+
+    # Setup Examples
+    gr.Examples(
+        examples=[
+            ["Hello! This is a test of the Supertonic 3 system running locally.", "M1", "English"],
+            ["こんにちは、これは日本語の音声合成のテストです。", "F1", "Japanese"],
+            ["¡Hola! Esta es una prueba de voz en español.", "F3", "Spanish"],
+            ["C'est un plaisir de vous rencontrer.", "M4", "French"]
+        ],
+        inputs=[input_text, voice_opt, lang_opt]
+    )
 
     btn.click(
         fn=generate_speech, 
         inputs=[input_text, voice_opt, lang_opt], 
-        outputs=[audio_output, stats]
+        outputs=[audio_output, status_box]
     )
 
 if __name__ == "__main__":