Update app.py

app.py

@@ -21,11 +21,15 @@ LICENSE = """
 ---.
 """
 
+models_cache = {}
+
 def load_model(model_id):
-    model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_4bit=True)
-    tokenizer = AutoTokenizer.from_pretrained(model_id)
-    tokenizer.use_default_system_prompt = False
-    return model, tokenizer
+    if model_id not in models_cache:
+        model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_4bit=True)
+        tokenizer = AutoTokenizer.from_pretrained(model_id)
+        tokenizer.use_default_system_prompt = False
+        models_cache[model_id] = (model, tokenizer)
+    return models_cache[model_id]
 
 if not torch.cuda.is_available():
     DESCRIPTION += "\n<p>Running on CPU 🥶 This demo does not work on CPU.</p>"
@@ -86,7 +90,7 @@ chat_interface = gr.ChatInterface(
     fn=generate,
     additional_inputs=[
         gr.Textbox(label="System prompt", lines=6),
-        gr.Textbox(label="Model ID", placeholder="Nekochu/Luminia-13B-v3"),
+        gr.Textbox(label="Model ID", default="Nekochu/Luminia-13B-v3"),
         gr.Slider(
             label="Max new tokens",
             minimum=1,
@@ -137,4 +141,4 @@ with gr.Blocks(css="style.css") as demo:
     gr.Markdown(LICENSE)
 
 if __name__ == "__main__":
-    demo.queue(max_size=20).launch()
+    demo.queue(max_size=20).launch()