Update app.py

app.py

@@ -26,8 +26,8 @@ max_seq_length=400
 attn_implementation=None
 
 #base_model = AutoModelForCausalLM.from_pretrained(model_name,return_dict=True,torch_dtype=torch.float16,)
-#base_model = AutoModelForCausalLM.from_pretrained(model_name,return_dict=True,device_map="auto", torch_dtype=torch.float16,)
-base_model = AutoModelForCausalLM.from_pretrained(base_model_name, return_dict=True, device_map = {"":0}, attn_implementation = attn_implementation,).eval()
+base_model = AutoModelForCausalLM.from_pretrained(base_model_name,return_dict=True,device_map="auto", torch_dtype=torch.float16,)
+#base_model = AutoModelForCausalLM.from_pretrained(base_model_name, return_dict=True, device_map = {"":0}, attn_implementation = attn_implementation,).eval()
 
 tokenizer = AutoTokenizer.from_pretrained(sft_model, max_length = max_seq_length)
 ft_model = PeftModel.from_pretrained(base_model, sft_model)