Update app.py

app.py

@@ -2,6 +2,7 @@ import gradio as gr
 import torch
 from TTS.api import TTS
 import os
+import time # Import time module for measuring performance
 
 # Agree to Coqui Terms of Service
 os.environ["COQUI_TOS_AGREED"] = "1"
@@ -10,13 +11,51 @@ os.environ["COQUI_TOS_AGREED"] = "1"
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 print(f"Running on: {device}")
 
+# --- Add check for model path before initialization ---
+model_path = os.path.expanduser("~/.local/share/tts/tts_models--multilingual--multi-dataset--xtts_v2")
+if not os.path.exists(model_path):
+    print(f"Model files not found at {model_path}. Initial download will occur.")
+else:
+    print(f"Model files found at {model_path}. Skipping initial download.")
+
+
 # Initialize TTS model (XTTS v2)
-tts = TTS(model_name="tts_models/multilingual/multi-dataset/xtts_v2", progress_bar=False, gpu=torch.cuda.is_available())
+print("Initializing TTS model...")
+try:
+    tts = TTS(model_name="tts_models/multilingual/multi-dataset/xtts_v2", progress_bar=False, gpu=torch.cuda.is_available())
+    print("TTS model initialized successfully.")
+except Exception as e:
+    print(f"Error initializing TTS model: {e}")
+    print("This might indicate issues with model files or environment setup.")
+    exit() # Exit if model fails to initialize
 
 # Define clone function
 def clone(text, audio):
-    tts.tts_to_file(text=text, speaker_wav=audio, language="en", file_path="./output.wav")
-    return "./output.wav"
+    print(f"Received request: Text='{text[:50]}...' (length {len(text)}), Audio='{audio}'") # Log input
+    
+    if not os.path.exists(audio):
+        print(f"Error: Reference audio file not found at {audio}")
+        gr.Warning("Reference audio file not found. Please ensure the path is correct.")
+        return None # Return None or handle error appropriately
+
+    start_time = time.time()
+    try:
+        print("Starting audio generation...")
+        tts.tts_to_file(text=text, speaker_wav=audio, language="en", file_path="./output.wav")
+        end_time = time.time()
+        print(f"Audio generated in {end_time - start_time:.2f} seconds.")
+        
+        if os.path.exists("./output.wav") and os.path.getsize("./output.wav") > 0:
+            print("Output file './output.wav' created successfully.")
+            return "./output.wav"
+        else:
+            print("Warning: Output file is empty or not created.")
+            gr.Warning("Audio generation completed, but output file might be empty or not created.")
+            return None # Indicate failure to Gradio
+    except Exception as e:
+        print(f"Error during audio generation: {e}")
+        gr.Error(f"An error occurred during audio generation: {e}")
+        return None # Indicate failure to Gradio
 
 # Build the Gradio Interface
 iface = gr.Interface(
@@ -29,9 +68,7 @@ iface = gr.Interface(
     title='Voice Clone',
     description="""
     by [Tony Assi](https://www.tonyassi.com/)
-
     This space uses the `xtts_v2` model. **Non-commercial use only**. [Coqui Public Model License](https://coqui.ai/cpml)
-
     Please ❤️ this Space. <a href="mailto:[email protected]">Email me</a>.
     """,
     theme=gr.themes.Base(primary_hue="teal", secondary_hue="teal", neutral_hue="slate"),
@@ -44,7 +81,9 @@ iface = gr.Interface(
         ["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"]
     ],
-    cache_examples=False  # <--- disables caching to avoid FileNotFoundError
+    cache_examples=False
 )
 
+print("\nLaunching Gradio interface...")
 iface.launch()
+print("Gradio interface launched.")