Cannot restart training after training tenc 2 AND using fused_backward_pass

If you finetune SDXL base with:
```
--train_text_encoder --learning_rate_te1 1e-10 --learning_rate_te2 1e-10 --fused_backward_pass
```
Then it will train fine. But if you stop training and restart by training from the e.g. `-step00001000.safetensors` file, you get this error message:
```
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [1280, 1280]], which is output 0 of AsStridedBackward0, is at version 1; expected version 0 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).
```
This doesn't happen if you only train te1 and the unet. It also only happens when you use --fused_backward_pass.

Full call stack:
```
Traceback (most recent call last):
File "/home/ara/m.2/Dev/sdxl/sd-scripts/./sdxl_train.py", line 944, in
train(args)
File "/home/ara/m.2/Dev/sdxl/sd-scripts/./sdxl_train.py", line 733, in train
accelerator.backward(loss)
File "/home/ara/m.2/Dev/sdxl/sd-scripts/venv/lib/python3.10/site-packages/accelerate/accelerator.py", line 1905, in backward
loss.backward(**kwargs)
File "/home/ara/m.2/Dev/sdxl/sd-scripts/venv/lib/python3.10/site-packages/torch/_tensor.py", line 522, in backward
torch.autograd.backward(
File "/home/ara/m.2/Dev/sdxl/sd-scripts/venv/lib/python3.10/site-packages/torch/autograd/__init__.py", line 266, in backward
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
File "/home/ara/m.2/Dev/sdxl/sd-scripts/venv/lib/python3.10/site-packages/torch/autograd/function.py", line 289, in apply
return user_fn(self, *args)
File "/home/ara/m.2/Dev/sdxl/sd-scripts/venv/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 319, in backward
torch.autograd.backward(outputs_with_grad, args_with_grad)
File "/home/ara/m.2/Dev/sdxl/sd-scripts/venv/lib/python3.10/site-packages/torch/autograd/__init__.py", line 266, in backward
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
```
(I also mentioned this bug back when the original pull request occurred: see https://github.com/kohya-ss/sd-scripts/pull/1259)

Reproduce the SDXL fine-tuning command with --train_text_encoder, both text encoders, and --fused_backward_pass, then restart from the reported step00001000.safetensors checkpoint. Start at sdxl_train.py:733 where accelerator.backward(loss) reaches the reported PyTorch checkpoint traceback, and review pull request #1259. Done means restarting from the checkpoint completes training without the in-place-operation error.

Written by the indexing model from the issue text.