When testing on the tabula sapiens, I ran with a heart dataset, then in the same session changed to a pancreas dataset. It seems when changing the dataset (rather than just the run) the metrics etc are still stored as the previous metrics:
which doesn't really make sense. Can we fix it so if the dataset is re-uploaded or re-selected (rather than just parameters being changed) we don't stash the previous results?
When testing on the tabula sapiens, I ran with a heart dataset, then in the same session changed to a pancreas dataset. It seems when changing the dataset (rather than just the run) the metrics etc are still stored as the previous metrics:
which doesn't really make sense. Can we fix it so if the dataset is re-uploaded or re-selected (rather than just parameters being changed) we don't stash the previous results?