Update app.py

app.py

@@ -1,40 +1,41 @@
-from flask import Flask, request, send_file, jsonify
+import spaces
+import gradio as gr
+import torch
 from TTS.api import TTS
 import os
-import tempfile
-
-# Accept TOS
 os.environ["COQUI_TOS_AGREED"] = "1"
 
-# Initialize TTS model once
-tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2").to("cpu")
-
-app = Flask(__name__)
-
-@app.route("/clone", methods=["POST"])
-def clone_route():
-    try:
-        # Get text and audio file from request
-        text = request.form.get("text")
-        audio_file = request.files.get("audio")
-
-        if not text or not audio_file:
-            return jsonify({"error": "Missing text or audio file"}), 400
-
-        # Save uploaded audio temporarily
-        with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_audio:
-            audio_path = tmp_audio.name
-            audio_file.save(audio_path)
-
-        # Generate cloned voice
-        output_path = tempfile.NamedTemporaryFile(delete=False, suffix=".wav").name
-        tts.tts_to_file(text=text, speaker_wav=audio_path, language="en", file_path=output_path)
-
-        # Return generated audio
-        return send_file(output_path, mimetype="audio/wav", as_attachment=True, download_name="output.wav")
-
-    except Exception as e:
-        return jsonify({"error": str(e)}), 500
-
-if __name__ == "__main__":
-    app.run(host="0.0.0.0", port=5000, debug=True)
+device = "cuda"
+
+tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2").to(device)
+
+@spaces.GPU(enable_queue=True)
+def clone(text, audio):
+    tts.tts_to_file(text=text, speaker_wav=audio, language="en", file_path="./output.wav")
+    return "./output.wav"
+
+iface = gr.Interface(fn=clone, 
+                     inputs=[gr.Textbox(label='Text'),gr.Audio(type='filepath', label='Voice reference audio file')], 
+                     outputs=gr.Audio(type='filepath'),
+                     title='Voice Clone',
+                     description="""
+                     by [Tony Assi](https://www.tonyassi.com/)
+
+                     ---
+
+                     ### If you like voice clone then try [Video Face Swap](https://huggingface.co/spaces/tonyassi/video-face-swap)! 
+
+                     ---
+                    
+                     This space uses xtts_v2 model. Non-commercial use only. [Coqui Public Model License](https://coqui.ai/cpml)
+                     
+                     Please ❤️ this Space. <a href="mailto: tony.assi.media@gmail.com">Email me</a>.
+                     """,
+                     theme = gr.themes.Base(primary_hue="teal",secondary_hue="teal",neutral_hue="slate"),
+                     examples=[["Hey! It's me Dorthy, from the Wizard of Oz. Type in whatever you'd like me to say.","./audio/Wizard-of-Oz-Dorthy.wav"],
+                               ["It's me Vito Corleone, from the Godfather. Type in whatever you'd like me to say.","./audio/Godfather.wav"],
+                               ["Hey, it's me Paris Hilton. Type in whatever you'd like me to say.","./audio/Paris-Hilton.mp3"],
+                               ["Hey, it's me Megan Fox from Transformers. Type in whatever you'd like me to say.","./audio/Megan-Fox.mp3"],
+                               ["Hey there, it's me Jeff Goldblum. Type in whatever you'd like me to say.","./audio/Jeff-Goldblum.mp3"],
+                               ["Hey there, it's me Heath Ledger as the Joker. Type in whatever you'd like me to say.","./audio/Heath-Ledger.mp3"],])
+iface.launch()