It appears when using NBEATs model while Generic_architecture is set to False that there is a tensor assigned to CPU while the rest is GPU:
RuntimeError Traceback (most recent call last)
in
7 model_name='nbeats_run_interp')
8
----> 9 nbeatsModel.fit(rescaled_train, val_series = rescaled_test, verbose = True)
~/.local/lib/python3.8/site-packages/darts/utils/torch.py in decorator(self, *args, **kwargs)
63 with fork_rng():
64 manual_seed(self._random_instance.randint(0, high=MAX_TORCH_SEED_VALUE))
---> 65 decorated(self, *args, **kwargs)
66 return decorator
~/.local/lib/python3.8/site-packages/darts/models/torch_forecasting_model.py in fit(self, series, covariates, val_series, val_covariates, verbose)
292 logger.info('Train dataset contains {} samples.'.format(len(train_dataset)))
293
--> 294 self.fit_from_dataset(train_dataset, val_dataset, verbose)
295
296 @random_method
~/.local/lib/python3.8/site-packages/darts/utils/torch.py in decorator(self, *args, **kwargs)
63 with fork_rng():
64 manual_seed(self._random_instance.randint(0, high=MAX_TORCH_SEED_VALUE))
---> 65 decorated(self, *args, **kwargs)
66 return decorator
~/.local/lib/python3.8/site-packages/darts/models/torch_forecasting_model.py in fit_from_dataset(self, train_dataset, val_dataset, verbose)
345
346 # Train model
--> 347 self._train(train_loader, val_loader, tb_writer, verbose)
348
349 # Close tensorboard writer
~/.local/lib/python3.8/site-packages/darts/models/torch_forecasting_model.py in _train(self, train_loader, val_loader, tb_writer, verbose)
499 self.model.train()
500 data, target = data.to(self.device), target.to(self.device)
--> 501 output = self.model(data)
502 loss = self.criterion(output, target)
503 self.optimizer.zero_grad()
~/.local/lib/python3.8/site-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
725 result = self._slow_forward(*input, **kwargs)
726 else:
--> 727 result = self.forward(*input, **kwargs)
728 for hook in itertools.chain(
729 _global_forward_hooks.values(),
~/.local/lib/python3.8/site-packages/darts/models/nbeats.py in forward(self, x)
306 for stack in self.stacks_list:
307 # compute stack output
--> 308 stack_residual, stack_forecast = stack(x)
309
310 # add stack forecast to final output
~/.local/lib/python3.8/site-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
725 result = self._slow_forward(*input, **kwargs)
726 else:
--> 727 result = self.forward(*input, **kwargs)
728 for hook in itertools.chain(
729 _global_forward_hooks.values(),
~/.local/lib/python3.8/site-packages/darts/models/nbeats.py in forward(self, x)
210 for block in self.blocks_list:
211 # pass input through block
--> 212 x_hat, y_hat = block(x)
213
214 # add block forecast to stack forecast
~/.local/lib/python3.8/site-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
725 result = self._slow_forward(*input, **kwargs)
726 else:
--> 727 result = self.forward(*input, **kwargs)
728 for hook in itertools.chain(
729 _global_forward_hooks.values(),
~/.local/lib/python3.8/site-packages/darts/models/nbeats.py in forward(self, x)
137
138 # waveform generator applications
--> 139 x_hat = self.backcast_g(theta_backcast)
140 y_hat = self.forecast_g(theta_forecast)
141
~/.local/lib/python3.8/site-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
725 result = self._slow_forward(*input, **kwargs)
726 else:
--> 727 result = self.forward(*input, **kwargs)
728 for hook in itertools.chain(
729 _global_forward_hooks.values(),
~/.local/lib/python3.8/site-packages/darts/models/nbeats.py in forward(self, x)
36
37 def forward(self, x):
---> 38 return torch.matmul(x, self.T.float().T)
39
40
RuntimeError: Tensor for argument #3 'mat2' is on CPU, but expected it to be on GPU (while checking arguments for addmm)
hrzn
commented
3 years ago
It appears when using NBEATs model while Generic_architecture is set to False that there is a tensor assigned to CPU while the rest is GPU: RuntimeError Traceback (most recent call last)