Update app.py

app.py

@@ -1,22 +1,28 @@
+import os
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import torch
 import json
 from datetime import datetime
 
+# Set Hugging Face Token for Authentication (ensure it's set in your environment)
+HUGGINGFACE_TOKEN = os.getenv("HUGGINGFACE_TOKEN")
+
 # Load Llama 3.2 (QLoRA) Model on CPU
 MODEL_NAME = "meta-llama/Llama-3.2-1B-Instruct-QLORA_INT4_EO8"
-tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME, token=HUGGINGFACE_TOKEN)
 model = AutoModelForCausalLM.from_pretrained(
     MODEL_NAME,
+    token=HUGGINGFACE_TOKEN,
     device_map="cpu"  # Force CPU usage
 )
 
 # Load Llama Guard for content moderation on CPU
 LLAMA_GUARD_NAME = "meta-llama/Llama-Guard-3-1B-INT4"
-guard_tokenizer = AutoTokenizer.from_pretrained(LLAMA_GUARD_NAME)
+guard_tokenizer = AutoTokenizer.from_pretrained(LLAMA_GUARD_NAME, token=HUGGINGFACE_TOKEN)
 guard_model = AutoModelForCausalLM.from_pretrained(
     LLAMA_GUARD_NAME,
+    token=HUGGINGFACE_TOKEN,
     device_map="cpu"
 )