Usage of FCSiamConc

When I used FCSiamConc model on Levir dataset I got

‘FCSiamConc’ object has no attribute ‘_current_fx_name’ error.Have you any suggestion for that?

from torch.utils.data import DataLoader
from torchgeo.datasets import LEVIRCDPlus
from torchgeo.models import FCSiamConc
from pytorch_lightning import Trainer

dataset = LEVIRCDPlus(root="lev", split="train", download=True, checksum=True)
dataloader = DataLoader(dataset, batch_size=4, shuffle=True, num_workers=4)
for sample in dataloader:
    image = sample["image"]
    target = sample["mask"]
model=FCSiamConc(classes=2)

trainer = Trainer(
    gpus=1,
    max_epochs=30,
    progress_bar_refresh_rate=30,
)

trainer.fit(model, dataloader)

_AttributeError                            Traceback (most recent call last)
Input In [1], in <cell line: 19>()
     11 model=FCSiamConc(classes=2)
     13 trainer = Trainer(
     14     gpus=1,
     15     max_epochs=30,
     16     progress_bar_refresh_rate=30,
     17 )
---> 19 trainer.fit(model, dataloader)

File ~\anaconda3\envs\thesis\lib\site-packages\pytorch_lightning\trainer\trainer.py:740, in Trainer.fit(self, model, train_dataloaders, val_dataloaders, datamodule, train_dataloader, ckpt_path)
    735     rank_zero_deprecation(
    736         "`trainer.fit(train_dataloader)` is deprecated in v1.4 and will be removed in v1.6."
    737         " Use `trainer.fit(train_dataloaders)` instead. HINT: added 's'"
    738     )
    739     train_dataloaders = train_dataloader
--> 740 self._call_and_handle_interrupt(
    741     self._fit_impl, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path
    742 )

File ~\anaconda3\envs\thesis\lib\site-packages\pytorch_lightning\trainer\trainer.py:685, in Trainer._call_and_handle_interrupt(self, trainer_fn, *args, **kwargs)
    675 r"""
    676 Error handling, intended to be used only for main trainer function entry points (fit, validate, test, predict)
    677 as all errors should funnel through them
   (...)
    682     **kwargs: keyword arguments to be passed to `trainer_fn`
    683 """
    684 try:
--> 685     return trainer_fn(*args, **kwargs)
    686 # TODO: treat KeyboardInterrupt as BaseException (delete the code below) in v1.7
    687 except KeyboardInterrupt as exception:

File ~\anaconda3\envs\thesis\lib\site-packages\pytorch_lightning\trainer\trainer.py:777, in Trainer._fit_impl(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)
    775 # TODO: ckpt_path only in v1.7
    776 ckpt_path = ckpt_path or self.resume_from_checkpoint
--> 777 self._run(model, ckpt_path=ckpt_path)
    779 assert self.state.stopped
    780 self.training = False

File ~\anaconda3\envs\thesis\lib\site-packages\pytorch_lightning\trainer\trainer.py:1125, in Trainer._run(self, model, ckpt_path)
   1122 # attach model to the training type plugin
   1123 self.training_type_plugin.connect(model)
-> 1125 self._callback_connector._attach_model_callbacks()
   1126 self._callback_connector._attach_model_logging_functions()
   1128 verify_loop_configurations(self)

File ~\anaconda3\envs\thesis\lib\site-packages\pytorch_lightning\trainer\connectors\callback_connector.py:266, in CallbackConnector._attach_model_callbacks(self)
    258 def _attach_model_callbacks(self) -> None:
    259     """Attaches the callbacks defined in the model.
    260 
    261     If a callback returned by the model's configure_callback method has the same type as one or several
   (...)
    264     will be pushed to the end of the list, ensuring they run last.
    265     """
--> 266     model_callbacks = self.trainer.call_hook("configure_callbacks")
    267     if not model_callbacks:
    268         return

File ~\anaconda3\envs\thesis\lib\site-packages\pytorch_lightning\trainer\trainer.py:1486, in Trainer.call_hook(self, hook_name, pl_module, *args, **kwargs)
   1484 pl_module = self.lightning_module or pl_module
   1485 if pl_module:
-> 1486     prev_fx_name = pl_module._current_fx_name
   1487     pl_module._current_fx_name = hook_name
   1489 # always profile hooks

File ~\anaconda3\envs\thesis\lib\site-packages\torch\nn\modules\module.py:1177, in Module.__getattr__(self, name)
   1175     if name in modules:
   1176         return modules[name]
-> 1177 raise AttributeError("'{}' object has no attribute '{}'".format(
   1178     type(self).__name__, name))

AttributeError: 'FCSiamConc' object has no attribute '_current_fx_name'_

[calebrob6]

torchgeo.models.FCSiamConc isn’t a LightningModule

[isaaccorley]

I think we should have a ChangeDetectionTask which I think would solve the confusion here since we can’t currently use FCSiam models with our current SegmentationTask.