Update app.py

app.py

@@ -3,8 +3,7 @@ from TTS.api import TTS
 from TTS.tts.configs.vits_config import VitsConfig
 from huggingface_hub import hf_hub_download
 import os
-import json
-import gc 
+import gc
 from romanizer import sinhala_to_roman 
 
 def load_my_model(repo_id):
@@ -21,56 +20,60 @@ def load_my_model(repo_id):
         if hasattr(config, "model_args"):
             config.model_args.num_chars = 137
     
-    # Initialize TTS without the 'items_per_group' argument
+    # Initialize TTS shell
     tts = TTS(gpu=False)
     
-    # Load the model using the config object we just fixed
-    tts.load_tts_model_by_path(
-        checkpoint_path=model_path,
-        config_path=config
-    )
+    # IMPORTANT: We pass model_path and config POSITIONALLY.
+    # Some TTS versions use 'checkpoint_path', others use 'model_path'.
+    # Passing them in order (1st, 2nd) avoids that 'unexpected keyword' error.
+    tts.load_tts_model_by_path(model_path, config)
     
     gc.collect()
     return tts
 
-# Load All Models
+# --- Loading Models ---
 print("Initializing Models...")
 try:
-    eng_tts = load_my_model("E-motionAssistant/text-to-speech-VITS-english")
-    sin_tts = load_my_model("E-motionAssistant/text-to-speech-VITS-sinhala")
-    tam_tts = load_my_model("E-motionAssistant/text-to-speech-VITS-tamil")
+    # We load them into a dictionary for cleaner access
+    engines = {
+        "English": load_my_model("E-motionAssistant/text-to-speech-VITS-english"),
+        "Sinhala": load_my_model("E-motionAssistant/text-to-speech-VITS-sinhala"),
+        "Tamil": load_my_model("E-motionAssistant/text-to-speech-VITS-tamil")
+    }
 except Exception as e:
     print(f"CRITICAL ERROR DURING LOADING: {e}")
+    engines = {}
 
 def generate_voice(text, language):
     try:
-        if language == "English":
-            engine = eng_tts
-            processed_text = text
-        elif language == "Sinhala":
-            engine = sin_tts
-            # Using your romanizer logic
+        engine = engines.get(language)
+        if engine is None:
+            return None
+
+        processed_text = text
+        if language == "Sinhala":
+            # Call your romanizer
             processed_text = sinhala_to_roman(text)
-        else:
-            engine = tam_tts
-            processed_text = text
+            print(f"Romanized: {processed_text}")
 
-        output_path = "output.wav"
-        engine.tts_to_file(text=processed_text, file_path=output_path)
+        # Unique output name to prevent file conflicts
+        output_path = f"output_{language.lower()}.wav"
+        engine.tts_to_file(text=str(processed_text), file_path=output_path)
         return output_path
     except Exception as e:
-        print(f"Error: {e}")
+        print(f"Inference Error: {e}")
         return None
 
 # Gradio Interface
 demo = gr.Interface(
     fn=generate_voice,
     inputs=[
-        gr.Textbox(label="Input Text"),
+        gr.Textbox(label="Input Text", placeholder="Type here..."),
         gr.Dropdown(["English", "Sinhala", "Tamil"], label="Select Language")
     ],
     outputs=gr.Audio(label="Synthesized Speech", type="filepath"),
     title="Multilingual VITS TTS"
 )
 
-demo.launch()
+if __name__ == "__main__":
+    demo.launch()