update app.py

app.py

@@ -9,6 +9,7 @@ import gradio as gr
 model_name = "Fredithefish/Guanaco-3B-Uncensored-v2"
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 model = AutoModelForCausalLM.from_pretrained(model_name)
+
 device = "cuda" if torch.cuda.is_available() else "cpu"
 model.to(device)
 
@@ -22,7 +23,7 @@ chat_history = []
 # --------------------------
 def generate_response(prompt):
     global chat_history
-    # Combine previous messages
+    # Combine previous conversation
     context = ""
     for user_msg, ai_msg in chat_history:
         context += f"User: {user_msg}\nAI: {ai_msg}\n"
@@ -39,6 +40,38 @@ def generate_response(prompt):
     chat_history.append((prompt, response))
     return response
 
+# --------------------------
+# Live typing effect
+# --------------------------
+def live_typing(prompt):
+    response = generate_response(prompt)
+    displayed_text = ""
+    for char in response:
+        displayed_text += char
+        time.sleep(0.02)  # typing speed
+        yield displayed_text
+
+# --------------------------
+# Gradio UI
+# --------------------------
+with gr.Blocks() as demo:
+    gr.Markdown("## 🤖 Guanaco-3B Chatbot with Live Typing")
+
+    # User input
+    user_input = gr.Textbox(label="Your Message", placeholder="Type something...")
+
+    # Output box
+    output_box = gr.Textbox(label="AI is typing...", lines=8)
+
+    # Submit event
+    user_input.submit(live_typing, inputs=[user_input], outputs=[output_box])
+
+demo.launch()
+    )
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True).split("AI:")[-1].strip()
+    chat_history.append((prompt, response))
+    return response
+
 # --------------------------
 # Simulate live typing
 # --------------------------