fix: unable to cache pipelines, thanks gradio

app.py

@@ -5,10 +5,6 @@ models = {
     "MagicPrompt": "pszemraj/distilgpt2-magicprompt-SD",
     "Llama-SmolTalk-3.2-1B": "prithivMLmods/Llama-SmolTalk-3.2-1B-Instruct"
 }
-pipelines = {}
-
-for key, value in models.items():
-    pipelines[key] = pipeline("text-generation", model=value)
 
 def respond(
     message,
@@ -19,7 +15,7 @@ def respond(
     top_p: float,
     top_k: int
 ):
-    yield pipelines[model](
+    yield pipeline("text-generation", model=model)(
         message,
         max_new_tokens=max_new_tokens,
         do_sample=True,
@@ -34,8 +30,11 @@ For information on how to customize the ChatInterface, peruse the gradio docs: h
 """
 demo = gr.ChatInterface(
     respond,
+    title="Prompt Enhancer Test",
+    type="messages",
     additional_inputs=[
         gr.Radio(choices=[models.items()], value="pszemraj/distilgpt2-magicprompt-SD", type="value", label="Model"),
+        # gr.Textbox(value="Enhance the provided text so that it is more vibrant and detailed.", label="System prompt"),
         gr.Slider(minimum=8, maximum=128, value=64, step=8, label="Max new tokens"),
         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
         gr.Slider(