Update app.py

app.py

@@ -3,7 +3,7 @@ from transformers import AutoModelForCausalLM, AutoTokenizer,BitsAndBytesConfig
 from transformers.generation import GenerationConfig
 quantization_config = BitsAndBytesConfig(
             load_in_4bit=True,
-            bnb_4bit_quant_type='nf4',
+            bnb_4bit_quant_type='int8',
             bnb_4bit_compute_dtype=torch.bfloat16)
 # Note: The default behavior now has injection attack prevention off.
 tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen-7B-Chat", trust_remote_code=True)