A question for valid error

[Youyi7]

Hi, I trained the model using SceneFlow dataset, then finetune them using kitti2015. When I get the final finetune result using kitti2015 dataset, my"total 3-px error in val = 2.932", and in your introduction, you get 1.89 valid result, so does my result resonable?

I hope you can give me some suggestions, thanks!