question_paper

question_paper.py

@@ -1,5 +1,6 @@
 import transformers
 import torch
+import os
 
 from fastapi import FastAPI
 
@@ -9,10 +10,10 @@ from transformers import AutoTokenizer
 
 app = FastAPI()
 model = "meta-llama/Llama-2-70b"
-
+access_token = os.getenv("access_token")
 @app.get("/")
 def llama():
-    tokenizer = AutoTokenizer.from_pretrained(model)
+    tokenizer = AutoTokenizer.from_pretrained(model,token=access_token)
 
     pipeline = transformers.pipeline("text-generation" ,model=model ,torch_dtype=torch.float16 ,device_map="auto" , )