pass audio file to transcriber directly

app.py

@@ -6,13 +6,9 @@ import gradio as gr
 import os
 import logging
 
-from vocode.helpers import create_turn_based_microphone_input_and_speaker_output
-
 from vocode.turn_based.agent.chat_gpt_agent import ChatGPTAgent
 from vocode.turn_based.synthesizer import CoquiSynthesizer
 from vocode.turn_based.transcriber.whisper_transcriber import WhisperTranscriber
-from vocode.turn_based.turn_based_conversation import TurnBasedConversation
-
 system_prompt = """
 You are the evil lord Darth Vader. You answer all questions your subjects ask you, but do so with immense contempt and loathing.
 You may use any of the following lines or take inspiration from this conversational style where appropriate in your response.
@@ -69,42 +65,61 @@ logger = logging.getLogger(__name__)
 logger.setLevel(logging.DEBUG)
 
 
-def main():
-    (
-        microphone_input,
-        speaker_output,
-    ) = create_turn_based_microphone_input_and_speaker_output(use_default_devices=True)
+def main(input_audio):
+
+    try:
+        # Initialize WhisperTranscriber
+        transcriber = WhisperTranscriber(api_key=getenv("OPENAI_API_KEY"))
+    except Exception as e:
+        logger.error("Failed to initialize WhisperTranscriber: %s", e)
+        return None
 
-    conversation = TurnBasedConversation(
-        input_device=microphone_input,
-        output_device=speaker_output,
-        transcriber=WhisperTranscriber(api_key=getenv("OPENAI_API_KEY")),
-        agent=ChatGPTAgent(
+    try:
+        # Initialize ChatGPTAgent
+        agent = ChatGPTAgent(
             system_prompt=system_prompt,
             initial_message="What up",
             api_key=getenv("OPENAI_API_KEY"),
-        ),
-        synthesizer=CoquiSynthesizer(
-            CoquiSynthesizerConfig.from_output_device(
-                speaker_output,
-                tts_kwargs={
-                    "voice_id": os.getenv("COQUI_VOICE_ID"),
-                }
-            ),
+        )
+    except Exception as e:
+        logger.error("Failed to initialize ChatGPTAgent: %s", e)
+        return None
+
+    try:
+        # Initialize CoquiSynthesizer
+        synthesizer = CoquiSynthesizer(
+            voice_id=os.getenv("COQUI_VOICE_ID"),
             api_key=getenv("COQUI_API_KEY"),
-        ),
-        logger=logger,
-    )
+        )
+    except Exception as e:
+        logger.error("Failed to initialize CoquiSynthesizer: %s", e)
+        return None
+
     print("Starting conversation. Press Ctrl+C to exit.")
     while True:
         try:
-            input("Press enter to start recording...")
-            conversation.start_speech()
-            input("Press enter to end recording...")
-            conversation.end_speech_and_respond()
-        except KeyboardInterrupt:
+            # Transcribe the input_audio using WhisperTranscriber
+            transcript = transcriber.transcribe(input_audio)
+        except Exception as e:
+            logger.error("Failed to transcribe audio: %s", e)
+            break
+
+        try:
+            # Generate response using ChatGPTAgent
+            response = agent.generate_response(transcript)
+        except Exception as e:
+            logger.error("Failed to generate response: %s", e)
+            break
+
+        try:
+            # Synthesize the response into audio using CoquiSynthesizer
+            output_audio = synthesizer.synthesize(response)
+        except Exception as e:
+            logger.error("Failed to synthesize response: %s", e)
             break
 
+        return output_audio
+
 
 demo = gr.Interface(fn=main, inputs="audio", outputs="audio")
 demo.launch()