Update app.py

app.py

@@ -28,7 +28,7 @@ def mistral_model():
         tuple: A tuple containing the loaded model and tokenizer.
     """
     model_id = "/code/mistral/"
-    model = AutoModelForCausalLM.from_pretrained("/home/user/app/mistral/")
+    model = AutoModelForCausalLM.from_pretrained("/home/user/app/mistral/",device_map="auto")
     tokenizer = AutoTokenizer.from_pretrained("/home/user/app/mistral/")
 
     return model,tokenizer
@@ -58,7 +58,7 @@ def load_model_norm():
 def mistral_generated_response(msg_prompt, persona_desc_prompt):
     user_prompt = f'{msg_prompt} [/INST]'
     persona_prompt = f'{persona_desc_prompt} [/INST]'
-    prompt_template = f'''### [INST] Instruction:{persona_prompt} [INST] {user_prompt}'''
+    prompt_template = f'''[INST] Instruction:{persona_prompt} [INST] {user_prompt}'''
     
     encodeds = tokenizer.apply_chat_template(prompt_template, return_tensors="pt")