Update app.py

app.py

@@ -61,9 +61,10 @@ class ModelManager:
                     with open(cache_file, "rb") as f:
                         self.models[model_name] = pickle.load(f)
                 else:
-                    self.models[model_name] = Llama.from_pretrained(repo_id=model_config['repo_id'], filename=model_config['filename'], use_auth_token=HUGGINGFACE_TOKEN)
+                    llama_model = Llama.from_pretrained(repo_id=model_config['repo_id'], filename=model_config['filename'], use_auth_token=HUGGINGFACE_TOKEN)
+                    self.models[model_name] = llama_model
                     with open(cache_file, "wb") as f:
-                        pickle.dump(self.models[model_name], f)
+                        pickle.dump(llama_model, f)
             except Exception as e:
                 print(f"Error loading {model_name}: {e}")
                 self.models[model_name] = None
@@ -132,7 +133,6 @@ async def api_generate_multimodel(request: Request):
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-
 iface = gr.Interface(
     fn=process_message,
     inputs=gr.Textbox(lines=2, placeholder="Enter your message here..."),
@@ -144,4 +144,4 @@ iface = gr.Interface(
 
 if __name__ == "__main__":
     port = int(os.environ.get("PORT", 7860))
-    iface.launch(server_port=port)
+    iface.launch(server_port=port)