8th Architecture

OS/01/conversations/user.json

@@ -1 +1 @@
-[{"role": "user", "type": "message", "content": "Hi, please just immediately run print hello world, don't do anything else.\n"}, {"role": "user", "type": "message", "content": "Can you calculate something like 1,933 times 2,345? Again, please just do it immediately. Don't do anything else. Just please run the code. Don't send a message to me.\n"}]
+[{"role": "user", "type": "message", "content": "Hi.\n"}]

OS/01/requirements.txt

@@ -5,6 +5,7 @@ pynput
 fastapi
 uvicorn
 websockets
+playsound
 python-dotenv
 ffmpeg-python
 textual

OS/01/start.sh

@@ -49,7 +49,7 @@ fi
 #     brew update
 #     brew install portaudio ffmpeg
 # fi
-# pip install -r requirements.txt
+python -m pip install -r requirements.txt
 
 ### START
 

OS/01/tts.py

@@ -6,6 +6,7 @@ import tempfile
 from openai import OpenAI
 from pydub import AudioSegment
 from pydub.playback import play
+from playsound import playsound
 
 client = OpenAI()
 
@@ -16,13 +17,10 @@ def tts(text, play_audio):
         input=text,
         response_format="mp3"
     )
-    with tempfile.NamedTemporaryFile() as temp_file:
+    with tempfile.NamedTemporaryFile(suffix=".mp3") as temp_file:
         response.stream_to_file(temp_file.name)
         
         if play_audio:
-            audio = AudioSegment.from_file(temp_file.name, format="mp3")
-            # Gradual fade in and out over 0.2 seconds
-            audio = audio.fade_in(200).fade_out(200)
-            play(audio)
+            playsound(temp_file.name)
         
         return temp_file.read()