both for linux and windows

app.py

@@ -3,8 +3,8 @@ import openai
 from decouple import config
 from gtts import gTTS
 import os
-import win32com.client
-import pythoncom
+import pydub
+import io
 import config 
 
 openai.api_key = config.API_KEYS['openai']
@@ -14,10 +14,6 @@ messages = [
     {"role": "system", "content": "You are a helpful assistant."},
 ]
 
-
-#  language = 'en'
-
-
 # Main method goes here
 def decipher(audio):
     global messages
@@ -34,14 +30,20 @@ def decipher(audio):
     )
 
     system_message = response["choices"][0]["message"]["content"]
-    pythoncom.CoInitialize()
-    speaker = win32com.client.Dispatch("SAPI.SpVoice")
-    speaker.Speak(system_message)
-    # myobj = gTTS(text=system_message, lang=language, slow=False)
-    # myobj.save("welcome.mp3")
-    # # Playing the converted file
-    # os.system("start welcome.mp3")
-    messages.append({"role": "assistant", "content": system_message},)
+    messages.append({"role": "assistant", "content": system_message})
+
+    # Convert the text to audio using gTTS
+    tts = gTTS(text=system_message, lang='en')
+    audio_data = io.BytesIO()
+    tts.write_to_fp(audio_data)
+
+    # Convert the audio to a playable format using pydub
+    audio_data.seek(0)
+    audio = pydub.AudioSegment.from_file(audio_data, format="mp3")
+
+    # Play the audio using VLC
+    player = pydub.playback.play
+    player(audio)
 
     chat_transcript = ""
     for message in messages:
@@ -50,8 +52,7 @@ def decipher(audio):
 
     return chat_transcript
 
-
 # Using Gradio's audio Interface 
 interface = gr.Interface(fn=decipher, inputs=gr.Audio(
     source="microphone", type="filepath"), outputs="text")
-interface.launch(share=True)
+interface.launch(share=True)