Update app.py

app.py

@@ -140,30 +140,6 @@ with gr.Blocks() as demo:
         Limits the number of tokens (words or subwords) the model can generate in its response. You can use this to control the length of the response.
         """)
 
-    gr.HTML(""" 
-    <style> 
-        #api_key_button { 
-            margin-top: 27px; 
-            background: linear-gradient(135deg, #4a00e0 0%, #8e2de2 100%); 
-        } 
-        #api_key_button:hover { 
-            background: linear-gradient(135deg, #5b10f1 0%, #9f3ef3 100%); 
-        } 
-        #clear_chat_button { 
-            background: linear-gradient(135deg, #e53e3e 0%, #f56565 100%); 
-        } 
-        #clear_chat_button:hover { 
-            background: linear-gradient(135deg, #c53030 0%, #e53e3e 100%); 
-        } 
-        #ask_button { 
-            background: linear-gradient(135deg, #fbd38d 0%, #f6e05e 100%); 
-        } 
-        #ask_button:hover { 
-            background: linear-gradient(135deg, #ecc94b 0%, #fbd38d 100%); 
-        } 
-    </style> 
-    """)
-
     # API Key Input
     with gr.Row():
         api_key_input = gr.Textbox(label="Enter OpenAI API Key", type="password")
@@ -180,30 +156,30 @@ with gr.Blocks() as demo:
             image_url = gr.Textbox(label="Enter Image URL")
             image_query = gr.Textbox(label="Ask about the Image")
             image_url_output = gr.Textbox(label="Response", interactive=False)
-            image_url_button = gr.Button("Ask",elem_id="ask_button")
+            image_url_button = gr.Button("Ask", elem_id="ask_button")
         
         with gr.Tab("Text Chat"):
             text_query = gr.Textbox(label="Enter your query")
             text_output = gr.Textbox(label="Response", interactive=False)
-            text_button = gr.Button("Ask",elem_id="ask_button")
+            text_button = gr.Button("Ask", elem_id="ask_button")
         
         with gr.Tab("Image Chat"):
             image_upload = gr.File(label="Upload an Image", type="filepath")
             image_text_query = gr.Textbox(label="Ask about the uploaded image")
             image_output = gr.Textbox(label="Response", interactive=False)
-            image_button = gr.Button("Ask",elem_id="ask_button")
+            image_button = gr.Button("Ask", elem_id="ask_button")
         
         with gr.Tab("PDF Chat"):
             pdf_upload = gr.File(label="Upload a PDF", type="filepath")
             pdf_text_query = gr.Textbox(label="Ask about the uploaded PDF")
             pdf_output = gr.Textbox(label="Response", interactive=False)
-            pdf_button = gr.Button("Ask",elem_id="ask_button")
-
-        with gr.Tab("Voice Chat"):
+            pdf_button = gr.Button("Ask", elem_id="ask_button")
+        
+        with gr.Tab("Voice Chat (Upload)"):
             audio_upload = gr.File(label="Upload an Audio File", type="binary")
             audio_query = gr.Textbox(label="Ask about the transcription")
             audio_output = gr.Textbox(label="Response", interactive=False)
-            audio_button = gr.Button("Ask",elem_id="ask_button")
+            audio_button = gr.Button("Ask", elem_id="ask_button")
 
         with gr.Tab("Voice(Record) Chat"):
             audio_record = gr.Audio(source="microphone", type="binary", label="Record your voice")
@@ -221,15 +197,15 @@ with gr.Blocks() as demo:
     image_button.click(image_chat, [image_upload, image_text_query, temperature, top_p, max_output_tokens], image_output)
     pdf_button.click(pdf_chat, [pdf_upload, pdf_text_query, temperature, top_p, max_output_tokens], pdf_output)
     
-    # For Voice Chat
+    # For Voice Chat (Upload)
     audio_button.click(
         lambda audio_binary, query, temperature, top_p, max_output_tokens: query_openai(
             [{"role": "user", "content": [{"type": "text", "text": transcribe_audio(audio_binary, api_key)}, {"type": "text", "text": query}]}],
             temperature, top_p, max_output_tokens
         ), [audio_upload, audio_query, temperature, top_p, max_output_tokens], audio_output
     )
-
-    # For Voice(Record) Chat
+    
+    # For Voice Chat (Record)
     audio_record_button.click(
         lambda audio_binary, query, temperature, top_p, max_output_tokens: query_openai(
             [{"role": "user", "content": [{"type": "text", "text": transcribe_audio(audio_binary, api_key)}, {"type": "text", "text": query}]}],