Update app.py

app.py

@@ -1,41 +1,33 @@
-import spaces
+import os
 import gradio as gr
-import torch
 from TTS.api import TTS
-import os
-os.environ["COQUI_TOS_AGREED"] = "1"
 
-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"
+# License Agreement
+os.environ["COQUI_TOS_AGREED"] = "1"
 
-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/)
+# Model Load (CPU Mode - GPU False)
+print("⏳ Model load ho raha hai... Isme 2-3 minute lagenge...")
+tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2", gpu=False)
 
-                     ---
+def predict(text, speaker_audio):
+    output_path = "output.wav"
+    # Hindi Generation
+    tts.tts_to_file(
+        text=text,
+        speaker_wav=speaker_audio,
+        language="hi",
+        file_path=output_path
+    )
+    return output_path
 
-                     ### If you like voice clone then try [Video Face Swap](https://huggingface.co/spaces/tonyassi/video-face-swap)! 
+# Simple Interface
+demo = gr.Interface(
+    fn=predict,
+    inputs=[
+        gr.Textbox(label="Text"),
+        gr.Audio(label="Reference Audio", type="filepath")
+    ],
+    outputs=gr.Audio(label="Generated Audio"),
+)
 
-                     ---
-                    
-                     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()
+demo.launch()