[Bug]: RuntimeError: none of output has requires_grad=True, this checkpoint() is not necessary

[hacknullchainik]

What happened?

Im finetunning sdxl and hitting 30 epoch i twice had faced this issue on 1st step 30 epoch

What did you expect would happen?

Id should't crash i guess

Relevant log output

step:   3%|█▋                                                           | 1/37 [00:19<11:42, 19.52s/it, loss=0.0329, smooth loss=0.0306]
epoch:  38%|████████████████████████████████▋                                                      | 30/80 [3:16:44<5:27:54, 393.49s/it]
Traceback (most recent call last):
  File "D:\Models\Training\OneTrainer\modules\ui\TrainUI.py", line 558, in __training_thread_function
    trainer.train()
  File "D:\Models\Training\OneTrainer\modules\trainer\GenericTrainer.py", line 662, in train
    loss.backward()
  File "D:\Models\Training\OneTrainer\venv\lib\site-packages\torch\_tensor.py", line 525, in backward
    torch.autograd.backward(
  File "D:\Models\Training\OneTrainer\venv\lib\site-packages\torch\autograd\__init__.py", line 267, in backward
    _engine_run_backward(
  File "D:\Models\Training\OneTrainer\venv\lib\site-packages\torch\autograd\graph.py", line 744, in _engine_run_backward
    return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
  File "D:\Models\Training\OneTrainer\venv\lib\site-packages\torch\autograd\function.py", line 301, in apply
    return user_fn(self, *args)
  File "D:\Models\Training\OneTrainer\venv\lib\site-packages\torch\utils\checkpoint.py", line 316, in backward
    raise RuntimeError(
RuntimeError: none of output has requires_grad=True, this checkpoint() is not necessary
Saving D:/Models/Training/OneFolders/dreamboth with florence med/models/result2222

Output of pip freeze

No response