Training phase sanity check fails by loading "../../data/imagenet/val" as an image

[wd255]

Original issue here: #17 Posting this new issue as the old one was marked completed and I cannot reopen.

Thanks @ShiFengyuan1999 for replying but I don't think the problem is solved. After replacing the path as instructed with , I still got similar error that raises when trying to load "/val" as image with PIL, but it's actually a directory. I think what the code is trying to do is reading whatever is under imagenet dir path as image. But as the dir structure you suggested

imagenet
└── train/
    ├── n01440764
        ├── n01440764_10026.JPEG
        ├── n01440764_10027.JPEG
        ├── ...
    ├── n01443537
    ├── ...
└── val/
    ├── ...

It's not the case. Am I structuring the folder in a wrong way?

---

Following the README.md but have the following error:

File "/home/duanwei/ML/Open-MAGVIT2/main.py", line 93, in [903/1986]
main()
File "/home/duanwei/ML/Open-MAGVIT2/main.py", line 87, in main
cli = LightningCLI(
^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/cli.py", line 388, in init
self._run_subcommand(self.subcommand)
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/cli.py", line 679, in _run_subcommand
fn(**fn_kwargs)
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/trainer/trainer.py", line 544, in fit
call._call_and_handle_interrupt(
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/trainer/call.py", line 43, in _call_and_handle_interrupt
return trainer.strategy.launcher.launch(trainer_fn, *args, trainer=trainer, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/strategies/launchers/subprocess_script.py", line 105, in launch
return function(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/trainer/trainer.py", line 580, in _fit_impl
self._run(model, ckpt_path=ckpt_path)
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/trainer/trainer.py", line 987, in _run
results = self._run_stage()
^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/trainer/trainer.py", line 1031, in _run_stage
self._run_sanity_check()
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/trainer/trainer.py", line 1060, in _run_sanity_check
val_loop.run()
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/loops/utilities.py", line 182, in _decorator
return loop_run(self, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/loops/evaluation_loop.py", line 128, in run
batch, batch_idx, dataloader_idx = next(data_fetcher)
^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/loops/fetchers.py", line 133, in next
batch = super().next()
^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/loops/fetchers.py", line 60, in next
batch = next(self.iterator)
^^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/utilities/combined_loader.py", line 341, in next
out = next(self._iterator)
^^^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/lightning/pytorch/utilities/combined_loader.py", line 142, in next
out = next(self.iterators[0])
^^^^^^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 634, in next
data = self._next_data()
^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 1346, in _next_data
return self._process_data(data)
^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 1372, in _process_data
data.reraise()
File "/root/anaconda3/lib/python3.11/site-packages/torch/_utils.py", line 644, in reraise
raise exception
IsADirectoryError: Caught IsADirectoryError in DataLoader worker process 0.

It seems that the sanity check code here
def _run_stage(self) -> Optional[Union[_PREDICT_OUTPUT, _EVALUATE_OUTPUT]]:
# wait for all to join if on distributed
self.strategy.barrier("run-stage")

zero_grad_kwargs = {} if _TORCH_GREATER_EQUAL_2_0 else {"set_to_none": True}
self.lightning_module.zero_grad(**zero_grad_kwargs)

if self.evaluating:
    return self._evaluation_loop.run()
if self.predicting:
    return self.predict_loop.run()
if self.training:
    with isolate_rng():
        # self._run_sanity_check()
    with torch.autograd.set_detect_anomaly(self._detect_anomaly):
        self.fit_loop.run()
    return None
raise RuntimeError(f"Unexpected state {self.state}")

I think it's caused by wrong data path setup.

For now I put the original ILSRVC directory of the kaggle version of imagenet under "../../data/" and renamed it to "imagenet". So that it has the structure as specified by README.md. Is this setting correct?
imagenet
└── train/
├── n01440764
├── n01440764_10026.JPEG
├── n01440764_10027.JPEG
├── ...
├── n01443537
├── ...
└── val/
├── ...