Update app.py

app.py

@@ -16,17 +16,15 @@ peft_model_id = "phearion/bigbrain-v0.0.1"
 config = PeftConfig.from_pretrained(peft_model_id)
 model = AutoModelForCausalLM.from_pretrained(
     config.base_model_name_or_path,                          
-    low_cpu_mem_usage=True,
-    return_dict=True,
     torch_dtype=torch.bfloat16)
 tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
 
 # Load the Lora model
 model = PeftModel.from_pretrained(model, peft_model_id)
-model = model.merge_and_unload()
 
 def greet(text):
-    batch = tokenizer(f"\"{text}\" ->: ", return_tensors='pt')
+    batch = tokenizer(f"\"{text
+}\" ->: ", return_tensors='pt')
     
     # Use torch.no_grad to disable gradient calculation
     with torch.no_grad():