Debasmita Ghose
Results
3
comments of
Debasmita Ghose
I have the same issue. The loss is not nan on the first iteration, but it gets to nan from second iteration onwards
Yes I did. Couple of things: Check the function `bbox_transform_batch(ex_rois, gt_rois)` in l`ib/model/rpn/bbox_transform.py`. It uses a logarithm to calculate `tagets_dw` and `targets_dh`. Check if these two variables are negative. If...
It works for modAL==0.3.6