Update app.py

app.py

@@ -20,7 +20,7 @@ model_gpt2 = GPT2LMHeadModel.from_pretrained('gpt2')
 model_name = "meta-llama/Meta-Llama-3.1-8B"
 try:
     model_llama = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto", torch_dtype=torch.float16,
-        load_in_8bit=True. token = hf_token)  # use device_map for automatic sharding
+        load_in_8bit=True, token = hf_token)  # use device_map for automatic sharding
 except OSError as e:
     print(f"Error loading model: {e}")