You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When trying the experiments, the decrease of the batch size would lead to the collapse of the training(The problem of DETR-Like Model). Better to keep or moderately decrease the size.
File "D:\code\MonoDETR\lib\models\monodetr\matcher.py", line 83, in forward
cost_giou = -generalized_box_iou(box_cxcylrtb_to_xyxy(out_bbox), box_cxcylrtb_to_xyxy(tgt_bbox))
File "D:\code\MonoDETR\utils\box_ops.py", line 62, in generalized_box_iou
assert (boxes1[:, 2:] >= boxes1[:, :2]).all()
AssertionError
epochs: 48%
I set batchsize to 8,GPU is 4060ti 16G
The text was updated successfully, but these errors were encountered: