Update app.py

app.py

@@ -47,10 +47,6 @@ model.eval()
 TTS_VOICES = [
     "en-US-JennyNeural",  # @tts1
     "en-US-GuyNeural",    # @tts2
-    "en-US-AriaNeural",   # @tts3
-    "en-US-DavisNeural",  # @tts4
-    "en-US-JaneNeural",   # @tts5
-    "en-US-JasonNeural",  # @tts6
 ]
 
 # Load multimodal (OCR) model and processor
@@ -105,8 +101,8 @@ def generate(
         images = []
 
     tts_prefix = "@tts"
-    is_tts = any(text.strip().lower().startswith(f"{tts_prefix}{i}") for i in range(1, 7))
-    voice_index = next((i for i in range(1, 7) if text.strip().lower().startswith(f"{tts_prefix}{i}")), None)
+    is_tts = any(text.strip().lower().startswith(f"{tts_prefix}{i}") for i in range(1, 3))
+    voice_index = next((i for i in range(1, 3) if text.strip().lower().startswith(f"{tts_prefix}{i}")), None)
     
     if is_tts and voice_index:
         voice = TTS_VOICES[voice_index - 1]