Update app.py

app.py

@@ -32,41 +32,46 @@ GREETING_MESSAGES = [
 def get_random_greeting():
     return random.choice(GREETING_MESSAGES)
 
-# Function to handle the chat response with streaming
 def respond_stream(message, history):
-    # Add the system message and previous chat history
-    system_message = "You are AstroSage, a highly knowledgeable AI assistant specialized in astronomy, astrophysics, and cosmology. Provide accurate, engaging, and educational responses about space science and the universe."
+    if not message:  # Handle empty messages
+        return
+
+    system_message = "You are AstroSage, a highly knowledgeable AI assistant..." # ... (your system message)
     messages = [{"role": "system", "content": system_message}]
-    for user_msg, assistant_msg in history:
-        if user_msg:
-            messages.append({"role": "user", "content": user_msg})
-        if assistant_msg:
-            messages.append({"role": "assistant", "content": assistant_msg})
+
+    # Format history correctly (especially important if you use clear)
+    for user, assistant in history:
+        messages.append({"role": "user", "content": user})
+        if assistant:  # Check if assistant message exists
+            messages.append({"role": "assistant", "content": assistant})
+
     messages.append({"role": "user", "content": message})
 
     try:
-        # Stream response from LLM
-        stream = llm.create_chat_completion(
+        response_content = ""
+        for chunk in llm.create_chat_completion(
             messages=messages,
             max_tokens=512,
             temperature=0.7,
             top_p=0.9,
-            stream=True  # Enable streaming
-        )
-        
-        # Stream the chunks of the response
-        response_content = ""
-        for chunk in stream:
-            response_content += chunk["choices"][0]["delta"]["content"]
-            yield response_content
+            stream=True
+        ):
+            delta = chunk["choices"][0]["delta"]
+            if "content" in delta: # check if content exists in delta
+                response_content += delta["content"]
+                yield response_content # yield inside the loop for streaming
     except Exception as e:
-        yield f"Error: {e}"
+        yield f"Error during generation: {e}"
+
 
-# Using gr.ChatInterface for a simpler chat UI
-chatbot = gr.ChatInterface(fn=respond_stream, type="messages")
 
-# Set a welcome message
+# Use gr.Chatbot for streaming and history management
+chatbot = gr.Chatbot()  # No need to specify fn here
 chatbot.set_welcome_message(get_random_greeting())
 
-if __name__ == "__main__":
-    chatbot.launch()
+with gr.Blocks() as demo:
+    chatbot.render()
+    clear = gr.Button("Clear")
+    clear.click(lambda: None, None, chatbot, fn=lambda: [])  # Correct clearing
+    chatbot.queue().launch()
+