Open Z-Lee-corder opened 7 months ago
Hello, have you solved this problem? I also meet this problem. That is ,when I train model on 3712 training split and obtain performance on 3769 validation split based on one model configuration. While when submitting to KITTI test server, the model configuration useful for validation does not work if I use 100% train +val split.
Hello, I want to ask, why did my kitti data set report the following errors during training.
Traceback (most recent call last):
File "detection/tools/train.py", line 204, in
Thank you for sharing this project!I have a issue to ask you: I try to train LoGoNet with all KITTI training data for 80 epochs and submit results to KITTI test set. But the precision of car on moderate level is only 82%, I want to know how to achieve 85.06% as paper said?
Is the model configuration on kitti test set different from that on kitti validation set?
If so, what changes need to be made?