Update app.py

app.py

@@ -24,7 +24,7 @@ model = PeftModel.from_pretrained(model, model_id)
 
 def greet(text):
     with torch.no_grad():  # Disable gradient calculation for inference
-        batch = tokenizer(f'### Input:\n{text}\n\n### Answer:\n', return_tensors='pt')  # Move tensors to device
+        batch = tokenizer(f'### Input:\n{text}\n\n### Answer:', return_tensors='pt')  # Move tensors to device
         with torch.cuda.amp.autocast():  # Enable mixed-precision if available
             output_tokens = model.generate(**batch, max_new_tokens=25)
     return tokenizer.decode(output_tokens[0], skip_special_tokens=True)