Update app_qwen_tts.py

app_qwen_tts.py

@@ -21,7 +21,6 @@ TTS_API_URL = "https://rahul7star-Chatterbox-Multilingual-TTS-API.hf.space/tts"
 # =========================================================
 BASE_DIR = os.path.dirname(os.path.abspath(__file__))
 DOC_PATH = os.path.join(BASE_DIR, DOC_FILE)
-
 if not os.path.exists(DOC_PATH):
     raise RuntimeError(f"❌ {DOC_FILE} not found next to app.py")
 
@@ -62,7 +61,7 @@ DOC_CHUNKS = chunk_text(DOC_TEXT)
 DOC_EMBEDS = embedder.encode(DOC_CHUNKS, normalize_embeddings=True, show_progress_bar=True)
 
 # =========================================================
-# Retrieve context
+# Retrieval
 # =========================================================
 def retrieve_context(question, k=TOP_K):
     q_emb = embedder.encode([question], normalize_embeddings=True)
@@ -83,7 +82,7 @@ def extract_final_answer(text: str) -> str:
     return lines[-1] if lines else text
 
 # =========================================================
-# Generate text answer
+# Generate text
 # =========================================================
 def answer_question(question):
     context = retrieve_context(question)
@@ -113,12 +112,13 @@ def answer_question(question):
     return extract_final_answer(decoded)
 
 # =========================================================
-# Call TTS API and get audio path
+# TTS API call
 # =========================================================
 def tts_via_api(text: str):
     try:
         payload = {"text": text}
-        resp = requests.post(TTS_API_URL, json=payload, timeout=60)
+        # no timeout, allow long requests
+        resp = requests.post(TTS_API_URL, json=payload)
         resp.raise_for_status()
         data = resp.json()
         audio_b64 = data.get("audio", "")
@@ -134,29 +134,26 @@ def tts_via_api(text: str):
         return None
 
 # =========================================================
-# Gradio Chat function
+# Gradio chat
 # =========================================================
 def chat(user_message, history):
     if not user_message.strip():
         return "", history
-
     try:
-        # Generate text
+        # 1️⃣ User text
+        history.append((user_message, None))  # None placeholder for bot column
+
+        # 2️⃣ Generate answer text
         answer_text = answer_question(user_message)
 
-        # Generate audio
+        # 3️⃣ Generate audio
         audio_path = tts_via_api(answer_text)
 
-        # Append tuple: (text, audio)
-        history.append((f"**Bot:** {answer_text}", audio_path))
-
-        # Append user message
-        history.append((f"**You:** {user_message}", None))
-
+        # 4️⃣ Append bot response
+        history.append((None, (answer_text, audio_path)))  # user=None, bot=(text, audio)
     except Exception as e:
         print(e)
-        history.append(("⚠️ Error generating response", None))
-
+        history.append((None, ("⚠️ Error generating response", None)))
     return "", history
 
 def reset_chat():
@@ -167,9 +164,11 @@ def reset_chat():
 # =========================================================
 def build_ui():
     with gr.Blocks(theme=gr.themes.Soft()) as demo:
-        gr.Markdown("# 📄 Qwen Document Assistant + TTS\nAsk questions and listen to answers.")
+        gr.Markdown("# 📄 Qwen Document Assistant + TTS\nAsk a question and listen to the answer.")
 
+        # Chat history: left=User, right=Bot
         chatbot = gr.Chatbot(height=450, type="tuples")
+
         msg = gr.Textbox(placeholder="Ask a question...", lines=2)
         send = gr.Button("Send")
         clear = gr.Button("🧹 Clear")