referring to the tutorial (https://pytorch-forecasting.readthedocs.io/en/stable/tutorials/stallion.html) provided by Pytorch about their implementation of the Temporal Fusion Transformer, I'm trying to use their calculate_prediction_actual_by_variable() and plot_prediction_actual_by_variable() methods on a custom dataset.
Although the "plot_prediction" method works correctly, the resulting plots using "plot_prediction_actual_by_variable()" do not show any 'prediction' or 'actual' curve. E.g:
Analyzing the tensors produced by the calculate_prediction_actual_by_variable() method, I realized that some NaNs values are present
Has anyone encountered this problem?
