Error when using CUDA

[Atai0610]

RuntimeError Traceback (most recent call last)
Cell In[29], line 2
1 trainer, model = build_trainer(rpu_config, log="mp")
----> 2 fit_model(trainer, model)
3 plot_loss(trainer, "Partial in-memory update using mixed precision.");

Cell In[22], line 26
23 def fit_model(trainer, model):
24 """Train the model using lightning trainer."""
---> 26 trainer.fit(model, *get_data_loaders(batch_size=5))
27 trainer.test(model, get_data_loaders(batch_size=100)[1])

File ~/miniconda3/envs/aihwkit/lib/python3.10/site-packages/pytorch_lightning/trainer/trainer.py:584, in Trainer.fit(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path, weights_only)
582 self.training = True
583 self.should_stop = False
--> 584 call._call_and_handle_interrupt(
585 self,
586 self._fit_impl,
587 model,
588 train_dataloaders,
589 val_dataloaders,
590 datamodule,
591 ckpt_path,
592 weights_only,
...
1207 x_input, d_input, self.analog_bias, self.in_trans, self.out_trans, self.non_blocking
1208 )

RuntimeError: CUDA_CALL Error 'the provided PTX was compiled with an unsupported toolchain.' at update_management_helper.cu:523
由于我的显卡是rtx5060 sm_120好像不支持 我该怎么操作

[PabloCarmona]

Hello @Atai0610 and thanks for using the toolkit! Can you give us little bit more of context on what you're doing when hitting this error? Share a examplo of the code, notebook, etc. you're trying to run and we will take a look on it 👀 Thanks!

[Atai0610]

trainer, model = build_trainer(rpu_config, log="mp") fit_model(trainer, model) plot_loss(trainer, "Partial in-memory update using mixed precision.");
官方的示例文档

[PabloCarmona]

Hey @Atai0610, is this error resolved after the guidance on the #775 discussion? Do you need any further help? Thanks!