Update app.py

app.py

@@ -1,4 +1,3 @@
-import os
 import base64
 import gradio as gr
 import openai
@@ -7,12 +6,6 @@ import io
 import tempfile
 import speech_recognition as sr
 
-# Initialize the OpenAI client
-client = openai.OpenAI(
-    base_url="https://llama3-2-3b.lepton.run/api/v1/",
-    api_key=os.environ.get('LEPTON_API_TOKEN')
-)
-
 def transcribe_audio(audio):
     # Convert the audio to wav format
     audio = AudioSegment.from_file(audio)
@@ -34,57 +27,73 @@ def transcribe_audio(audio):
 
     return text
 
-def process_audio(audio):
+def process_audio(audio, api_token):
+    if not api_token:
+        return "Please provide an API token.", None
+
+    # Initialize the OpenAI client with the user-provided token
+    client = openai.OpenAI(
+        base_url="https://llama3-2-3b.lepton.run/api/v1/",
+        api_key=api_token
+    )
+
     # Transcribe the input audio
     transcription = transcribe_audio(audio)
 
-    # Process the transcription with the API
-    completion = client.chat.completions.create(
-        model="gpt-3.5-turbo",
-        messages=[
-            {"role": "user", "content": transcription},
-        ],
-        max_tokens=128,
-        stream=True,
-        extra_body={
-            "require_audio": "true",
-            "tts_preset_id": "jessica",
-        }
-    )
+    try:
+        # Process the transcription with the API
+        completion = client.chat.completions.create(
+            model="gpt-3.5-turbo",
+            messages=[
+                {"role": "user", "content": transcription},
+            ],
+            max_tokens=128,
+            stream=True,
+            extra_body={
+                "require_audio": "true",
+                "tts_preset_id": "jessica",
+            }
+        )
 
-    response_text = ""
-    audios = []
+        response_text = ""
+        audios = []
 
-    for chunk in completion:
-        if not chunk.choices:
-            continue
-        content = chunk.choices[0].delta.content
-        audio = getattr(chunk.choices[0], 'audio', [])
-        if content:
-            response_text += content
-        if audio:
-            audios.extend(audio)
+        for chunk in completion:
+            if not chunk.choices:
+                continue
+            content = chunk.choices[0].delta.content
+            audio = getattr(chunk.choices[0], 'audio', [])
+            if content:
+                response_text += content
+            if audio:
+                audios.extend(audio)
 
-    # Combine audio chunks and save as MP3
-    audio_data = b''.join([base64.b64decode(audio) for audio in audios])
-    
-    # Save the audio to a temporary file
-    with tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) as temp_audio:
-        temp_audio.write(audio_data)
-        temp_audio_path = temp_audio.name
+        # Combine audio chunks and save as MP3
+        audio_data = b''.join([base64.b64decode(audio) for audio in audios])
+        
+        # Save the audio to a temporary file
+        with tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) as temp_audio:
+            temp_audio.write(audio_data)
+            temp_audio_path = temp_audio.name
 
-    return response_text, temp_audio_path
+        return response_text, temp_audio_path
+
+    except Exception as e:
+        return f"An error occurred: {str(e)}", None
 
 # Create the Gradio interface
 iface = gr.Interface(
     fn=process_audio,
-    inputs=gr.Audio(type="filepath"),
+    inputs=[
+        gr.Audio(type="filepath", label="Input Audio"),
+        gr.Textbox(label="API Token", type="password")
+    ],
     outputs=[
         gr.Textbox(label="Response Text"),
         gr.Audio(label="Response Audio")
     ],
     title="Audio-to-Audio Demo",
-    description="Upload an audio file to get a response in both text and audio format."
+    description="Upload an audio file and provide your API token to get a response in both text and audio format."
 )
 
 # Launch the interface