Update app.py

app.py

@@ -1,15 +1,11 @@
 import gradio as gr
 
-import torch
-from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
-from transformers import TextStreamer
-import transformers
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
 import spaces
-
-@spaces.GPU
-def load_pretrained_model():
-    return AutoModelForCausalLM.from_pretrained("Rorical/0-roleplay", return_dict=True, trust_remote_code=True)
+  
+tokenizer = AutoTokenizer.from_pretrained("Rorical/0-roleplay", trust_remote_code=True)
+tokenizer.chat_template = "{% for message in messages %}{{'<|im_start|>' + ((message['role'] + ':\n') if message['role'] != '' else '') + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>星野:\n' }}{% endif %}" # Be careful that this model used custom chat template.
 
 # Define the response function
 @spaces.GPU
@@ -21,8 +17,8 @@ def respond(
     temperature,
     top_p,
 ):
-    global model, tokenizer
-    
+    model = AutoModelForCausalLM.from_pretrained("Rorical/0-roleplay", return_dict=True, trust_remote_code=True)
+        
     # Construct the messages for the chat
     messages = [{"role": "", "content": system_message}]
     for user_message, bot_response in history:
@@ -83,10 +79,7 @@ demo = gr.ChatInterface(
 )
 
 if __name__ == "__main__":
-    global model, tokenizer
-    
-    tokenizer = AutoTokenizer.from_pretrained("Rorical/0-roleplay", trust_remote_code=True)
-    tokenizer.chat_template = "{% for message in messages %}{{'<|im_start|>' + ((message['role'] + ':\n') if message['role'] != '' else '') + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>星野:\n' }}{% endif %}" # Be careful that this model used custom chat template.
+    global model
 
     model = load_pretrained_model()