jdb78
commented
3 years ago You need to update pytorch forecasting. Should do the trick.
Closed jiryang closed 3 years ago
jdb78
commented
3 years ago You need to update pytorch forecasting. Should do the trick.
Expected behavior
Getting baseline prediction result (using SMAPE) of Nbeats sample code shows the below error: Traceback (most recent call last): File "welt_forecast_nbeats.py", line 90, in
print(SMAPE()(baseline_predictions, actuals))
File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 744, in _call_impl
result = self.forward(*input, kwargs)
File "/usr/local/lib/python3.6/dist-packages/pytorch_lightning/metrics/metric.py", line 153, in forward
self.update(*args, *kwargs)
File "/usr/local/lib/python3.6/dist-packages/pytorch_lightning/metrics/metric.py", line 199, in wrapped_func
return update(args, kwargs)
File "/usr/local/lib/python3.6/dist-packages/pytorch_forecasting/metrics.py", line 308, in update
target, lengths = unpack_sequence(target)
File "/usr/local/lib/python3.6/dist-packages/pytorch_forecasting/utils.py", line 232, in unpack_sequence
lengths = torch.ones(sequence.size(0), device=sequence.device, dtype=torch.long) * sequence.size(1)
IndexError: Dimension out of range (expected to be in range of [-1, 0], but got 1)
Code to reproduce the problem
calculate baseline absolute error
actuals = torch.cat([y[0] for x, y in iter(val_dataloader)]) baseline_predictions = Baseline().predict(val_dataloader) SMAPE()(baseline_predictions, actuals)