Update app.py

app.py

@@ -70,7 +70,7 @@ class WebSocketClient:
             if data["type"] == "conversation.item.input_audio_transcription.delta":
                 self.transcript += data["delta"]
 
-# WebSocket connection setup
+# Connection manager
 def create_ws():
     cid = str(uuid.uuid4())
     client = WebSocketClient(WS_URI, HEADERS, cid)
@@ -82,11 +82,8 @@ def send_audio(chunk, cid):
     if not cid or cid not in connections:
         return "Connecting..."
     sr, arr = chunk
-
-    # Reset transcript if it's been running long or restarted
     if len(connections[cid].transcript) > 1000:
         connections[cid].transcript = ""
-
     connections[cid].enqueue_audio_chunk(sr, arr)
     return connections[cid].transcript.strip()
 
@@ -95,7 +92,6 @@ def clear_transcript(cid):
         connections[cid].transcript = ""
     return ""
 
-# Chat assistant logic
 def handle_chat(user_input, history, thread_id, image_url):
     if not OPENAI_API_KEY or not ASSISTANT_ID:
         return "❌ Missing secrets!", history, thread_id, image_url
@@ -134,7 +130,6 @@ def handle_chat(user_input, history, thread_id, image_url):
 def send_transcript_to_assistant(transcript, history, thread_id, image_url, cid):
     if not transcript.strip():
         return gr.update(), history, thread_id, image_url
-    # Clear transcript after sending
     if cid in connections:
         connections[cid].transcript = ""
     return handle_chat(transcript, history, thread_id, image_url)
@@ -156,12 +151,24 @@ with gr.Blocks(theme=gr.themes.Soft()) as app:
         border-radius: 6px;
         margin-top: 10px;
         background-color: #f2f2f2 !important;
+        color: #000 !important;
+        white-space: nowrap;
+        overflow: hidden;
+        text-overflow: ellipsis;
     }
+
     button {
         margin-right: 8px;
     }
+
+    /* Hide icon (optional) */
     #record-audio button svg {
-        display: none !important;
+        margin-right: 6px;
+    }
+
+    /* Hide internal label if redundant */
+    #record-audio label {
+        display: none;
     }
     </style>
     """)
@@ -182,7 +189,8 @@ with gr.Blocks(theme=gr.themes.Soft()) as app:
                 send_btn = gr.Button("Send", variant="primary", scale=2)
 
             with gr.Accordion("🎤 Voice Transcription", open=False) as voice_section:
-                voice_input = gr.Audio(label="🎙️ Record", streaming=True, elem_id="record-audio")
+                gr.Markdown("**🎙️ Tap below to record your voice**")
+                voice_input = gr.Audio(label="", streaming=True, elem_id="record-audio")
                 voice_transcript = gr.Textbox(label="Transcript", lines=2, interactive=False)
 
                 with gr.Row():
@@ -194,7 +202,6 @@ with gr.Blocks(theme=gr.themes.Soft()) as app:
                    inputs=[user_prompt, chat_state, thread_state, image_state],
                    outputs=[user_prompt, chat, thread_state, image_state])
     image_state.change(fn=lambda x: x, inputs=image_state, outputs=image_display)
-
     voice_input.stream(fn=send_audio, inputs=[voice_input, client_id], outputs=voice_transcript, stream_every=0.5)
 
     ask_btn.click(fn=send_transcript_to_assistant,