I’m using MultiModalSupervisedDataset for data, SequenceToTarget for model initialization.
Getting this warning: UserWarning: The compute method of metric AUROC was called before the update method which may lead to errors, as metric states have not yet been updated.
warnings.warn(*args, **kwargs)
Can someone help ?
I'll be glad if you provide me demo for MultiModalSupervisedDataset, SequenceToTarget classes.
Full error log:
ValueError Traceback (most recent call last)
Input In [52], in <cell line: 2>()
1 print(f'logger.version = {pl_trainer.logger.version}')
----> 2 pl_trainer.fit(model, train_loader)
3 print(pl_trainer.logged_metrics)
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:770, in Trainer.fit(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)
751 r"""
752 Runs the full optimization routine.
753
(...)
767 datamodule: An instance of :class:~pytorch_lightning.core.datamodule.LightningDataModule.
768 """
769 self.strategy.model = model
--> 770 self._call_and_handle_interrupt(
771 self._fit_impl, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path
772 )
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:723, in Trainer._call_and_handle_interrupt(self, trainer_fn, *args, kwargs)
721 return self.strategy.launcher.launch(trainer_fn, *args, trainer=self, *kwargs)
722 else:
--> 723 return trainer_fn(args, kwargs)
724 # TODO: treat KeyboardInterrupt as BaseException (delete the code below) in v1.7
725 except KeyboardInterrupt as exception:
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:811, in Trainer._fit_impl(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)
807 ckpt_path = ckpt_path or self.resume_from_checkpoint
808 self._ckpt_path = self.__set_ckpt_path(
809 ckpt_path, model_provided=True, model_connected=self.lightning_module is not None
810 )
--> 811 results = self._run(model, ckpt_path=self.ckpt_path)
813 assert self.state.stopped
814 self.training = False
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1236, in Trainer._run(self, model, ckpt_path)
1233 self._checkpoint_connector.restore_training_state()
1235 self._checkpoint_connector.resume_end()
-> 1237 results = self._run_stage()
1238 log.detail(f"{self.class.name}: trainer tearing down")
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1323, in Trainer._run_stage(self)
1321 if self.predicting:
1322 return self._run_predict()
-> 1323 return self._run_train()
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1345, in Trainer._run_train(self)
1342 self._pre_training_routine()
1344 with isolate_rng():
-> 1345 self._run_sanity_check()
1347 # enable train mode
1348 self.model.train()
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1413, in Trainer._run_sanity_check(self)
1411 # run eval step
1412 with torch.no_grad():
-> 1413 val_loop.run()
1415 self._call_callback_hooks("on_sanity_check_end")
1417 # reset logger connector
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/loops/base.py:211, in Loop.run(self, *args, kwargs)
208 break
209 self._restarting = False
--> 211 output = self.on_run_end()
212 return output
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/loops/dataloader/evaluation_loop.py:183, in EvaluationLoop.on_run_end(self)
180 self.trainer._logger_connector.epoch_end_reached()
182 # hook
--> 183 self._evaluation_epoch_end(self._outputs)
184 self._outputs = [] # free memory
186 # hook
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/loops/dataloader/evaluation_loop.py:311, in EvaluationLoop._evaluation_epoch_end(self, outputs)
309 self.trainer._call_lightning_module_hook("test_epoch_end", output_or_outputs)
310 else:
--> 311 self.trainer._call_lightning_module_hook("validation_epoch_end", output_or_outputs)
File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1595, in Trainer._call_lightning_module_hook(self, hook_name, pl_module, *args, *kwargs)
1592 pl_module._current_fx_name = hook_name
1594 with self.profiler.profile(f"[LightningModule]{pl_module.class.name}.{hook_name}"):
-> 1595 output = fn(args, kwargs)
1597 # restore current_fx when nested context
1598 pl_module._current_fx_name = prev_fx_name
File ~/.local/lib/python3.9/site-packages/ptls/frames/supervised/seq_to_target.py:148, in SequenceToTarget..validation_epoch_end(self, outputs)
147 for name, mf in self.valid_metrics.items():
--> 148 self.log(f'valid/{name}', mf.compute(), prog_bar=True)
149 for name, mf in self.valid_metrics.items():
150 mf.reset()
File ~/.local/lib/python3.9/site-packages/torchmetrics/metric.py:531, in Metric._wrap_compute..wrapped_func(*args, *kwargs)
523 # compute relies on the sync context manager to gather the states across processes and apply reduction
524 # if synchronization happened, the current rank accumulated states will be restored to keep
525 # accumulation going if should_unsync=True,
526 with self.sync_context(
527 dist_sync_fn=self.dist_sync_fn, # type: ignore
528 should_sync=self._to_sync,
529 should_unsync=self._should_unsync,
530 ):
--> 531 value = compute(args, **kwargs)
532 self._computed = _squeeze_if_scalar(value)
534 return self._computed
File ~/.local/lib/python3.9/site-packages/torchmetrics/classification/auroc.py:172, in AUROC.compute(self)
171 raise RuntimeError("You have to have determined mode.")
--> 172 preds = dim_zero_cat(self.preds)
173 target = dim_zero_cat(self.target)
174 return _auroc_compute(
175 self.preds,
176 self.target,
(...)
181 self.max_fpr,
182 )
File ~/.local/lib/python3.9/site-packages/torchmetrics/utilities/data.py:41, in dim_zero_cat(x)
39 x = [y.unsqueeze(0) if y.numel() == 1 and y.ndim == 0 else y for y in x]
40 if not x: # empty list
---> 41 raise ValueError("No samples to concatenate")
42 return torch.cat(x, dim=0)
ValueError: No samples to concatenate
I’m using MultiModalSupervisedDataset for data, SequenceToTarget for model initialization. Getting this warning: UserWarning: The
computemethod of metric AUROC was called before theupdatemethod which may lead to errors, as metric states have not yet been updated. warnings.warn(*args, **kwargs) Can someone help ? I'll be glad if you provide me demo for MultiModalSupervisedDataset, SequenceToTarget classes.Full error log:
ValueError Traceback (most recent call last) Input In [52], in <cell line: 2>() 1 print(f'logger.version = {pl_trainer.logger.version}') ----> 2 pl_trainer.fit(model, train_loader) 3 print(pl_trainer.logged_metrics) File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:770, in Trainer.fit(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path) 751 r""" 752 Runs the full optimization routine. 753 (...) 767 datamodule: An instance of :class:.wrapped_func(*args, *kwargs)
523 # compute relies on the sync context manager to gather the states across processes and apply reduction
524 # if synchronization happened, the current rank accumulated states will be restored to keep
525 # accumulation going if
~pytorch_lightning.core.datamodule.LightningDataModule. 768 """ 769 self.strategy.model = model --> 770 self._call_and_handle_interrupt( 771 self._fit_impl, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path 772 ) File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:723, in Trainer._call_and_handle_interrupt(self, trainer_fn, *args, kwargs) 721 return self.strategy.launcher.launch(trainer_fn, *args, trainer=self, *kwargs) 722 else: --> 723 return trainer_fn(args, kwargs) 724 # TODO: treat KeyboardInterrupt as BaseException (delete the code below) in v1.7 725 except KeyboardInterrupt as exception: File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:811, in Trainer._fit_impl(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path) 807 ckpt_path = ckpt_path or self.resume_from_checkpoint 808 self._ckpt_path = self.__set_ckpt_path( 809 ckpt_path, model_provided=True, model_connected=self.lightning_module is not None 810 ) --> 811 results = self._run(model, ckpt_path=self.ckpt_path) 813 assert self.state.stopped 814 self.training = False File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1236, in Trainer._run(self, model, ckpt_path) 1233 self._checkpoint_connector.restore_training_state() 1235 self._checkpoint_connector.resume_end() -> 1237 results = self._run_stage() 1238 log.detail(f"{self.class.name}: trainer tearing down") File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1323, in Trainer._run_stage(self) 1321 if self.predicting: 1322 return self._run_predict() -> 1323 return self._run_train() File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1345, in Trainer._run_train(self) 1342 self._pre_training_routine() 1344 with isolate_rng(): -> 1345 self._run_sanity_check() 1347 # enable train mode 1348 self.model.train() File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1413, in Trainer._run_sanity_check(self) 1411 # run eval step 1412 with torch.no_grad(): -> 1413 val_loop.run() 1415 self._call_callback_hooks("on_sanity_check_end") 1417 # reset logger connector File ~/.local/lib/python3.9/site-packages/pytorch_lightning/loops/base.py:211, in Loop.run(self, *args, kwargs) 208 break 209 self._restarting = False --> 211 output = self.on_run_end() 212 return output File ~/.local/lib/python3.9/site-packages/pytorch_lightning/loops/dataloader/evaluation_loop.py:183, in EvaluationLoop.on_run_end(self) 180 self.trainer._logger_connector.epoch_end_reached() 182 # hook --> 183 self._evaluation_epoch_end(self._outputs) 184 self._outputs = [] # free memory 186 # hook File ~/.local/lib/python3.9/site-packages/pytorch_lightning/loops/dataloader/evaluation_loop.py:311, in EvaluationLoop._evaluation_epoch_end(self, outputs) 309 self.trainer._call_lightning_module_hook("test_epoch_end", output_or_outputs) 310 else: --> 311 self.trainer._call_lightning_module_hook("validation_epoch_end", output_or_outputs) File ~/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1595, in Trainer._call_lightning_module_hook(self, hook_name, pl_module, *args, *kwargs) 1592 pl_module._current_fx_name = hook_name 1594 with self.profiler.profile(f"[LightningModule]{pl_module.class.name}.{hook_name}"): -> 1595 output = fn(args, kwargs) 1597 # restore current_fx when nested context 1598 pl_module._current_fx_name = prev_fx_name File ~/.local/lib/python3.9/site-packages/ptls/frames/supervised/seq_to_target.py:148, in SequenceToTarget..validation_epoch_end(self, outputs) 147 for name, mf in self.valid_metrics.items(): --> 148 self.log(f'valid/{name}', mf.compute(), prog_bar=True) 149 for name, mf in self.valid_metrics.items(): 150 mf.reset() File ~/.local/lib/python3.9/site-packages/torchmetrics/metric.py:531, in Metric._wrap_compute.should_unsync=True, 526 with self.sync_context( 527 dist_sync_fn=self.dist_sync_fn, # type: ignore 528 should_sync=self._to_sync, 529 should_unsync=self._should_unsync, 530 ): --> 531 value = compute(args, **kwargs) 532 self._computed = _squeeze_if_scalar(value) 534 return self._computed File ~/.local/lib/python3.9/site-packages/torchmetrics/classification/auroc.py:172, in AUROC.compute(self) 171 raise RuntimeError("You have to have determined mode.") --> 172 preds = dim_zero_cat(self.preds) 173 target = dim_zero_cat(self.target) 174 return _auroc_compute( 175 self.preds, 176 self.target, (...) 181 self.max_fpr, 182 ) File ~/.local/lib/python3.9/site-packages/torchmetrics/utilities/data.py:41, in dim_zero_cat(x) 39 x = [y.unsqueeze(0) if y.numel() == 1 and y.ndim == 0 else y for y in x] 40 if not x: # empty list ---> 41 raise ValueError("No samples to concatenate") 42 return torch.cat(x, dim=0) ValueError: No samples to concatenate