added accelerate to the requirements

load_meta_data.py

@@ -16,6 +16,7 @@ class ChatBot:
         self.tokenizer =  AutoTokenizer.from_pretrained(model_id,
                                           device_map = 'auto')
         self.model = AutoModelForCausalLM.from_pretrained(model_id,
+                                                    low_cpu_mem_usage=True,
                                              device_map='auto')
         
     def get_response(self,text: UserQuery) -> str:

requirements.txt

@@ -1,4 +1,6 @@
 transformers
+accelerate 
+bitsandbytes
 fastapi
 torch
 uvicorn