There are still some issues with the results for FGC (and maybe FtGC), I think that is because of the data. Either way, it should not effect other functionality, including training and testing/evaluating on the remaining tasks.
I merged our implementations for display_labels_on_images.py to include all of our tasks.
General fixes for all scripts, the following were tested and are now working:
There are still some issues with the results for FGC (and maybe FtGC), I think that is because of the data. Either way, it should not effect other functionality, including training and testing/evaluating on the remaining tasks.
I merged our implementations for
display_labels_on_images.py
to include all of our tasks.