Update app.py

app.py

@@ -1,15 +1,17 @@
+import os
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import torch
 
 model_id = "hugging-quants/Meta-Llama-3.1-8B-Instruct-GPTQ-INT4"
+hf_token = os.environ.get("HF_TOKEN")
 
-tokenizer = AutoTokenizer.from_pretrained(model_id, use_auth_token=True)
+tokenizer = AutoTokenizer.from_pretrained(model_id, token=hf_token)
 model = AutoModelForCausalLM.from_pretrained(
     model_id,
     device_map="auto",
     torch_dtype=torch.float16,
     low_cpu_mem_usage=True,
-    use_auth_token=True  # ⬅️ ensures gated model access
+    token=hf_token
 )
 
 def generate(prompt):