Validation vs. Test set for the burn scars dataset

[gabrieltseng]

Hello!

Thank you for releasing these datasets.

I understand that you use the "validation set" to report test results in the paper. It also seems like you use this set as a "true" validation set:

hls-foundation-os/configs/burn_scars.py

Line 141 in df7761e

data = dict(

Did you use the "val/test" set for both validation / hyperparameter selection and reporting test results?

If not, what split did you use for validation & hyperparameter selection?

Thanks!

[gabrieltseng]

In addition, would it be possible to share the actual values (for all metrics, in addition to IoU for the wildfire scar class) used to make the plots in Table 10 of the paper?