Exact training setup

[vobecant]

Dear authors,

thank you very much for providing us the code for the "KM3D" paper (is it code corresponding to KM3D and not to RTM3D, right?).

I would like to reproduce your results on the KITTI dataset and then apply your method to the Waymo Open dataset. Therefore, I would like to know your exact training setup. Namely, I am interested in:

• batch size
• number of GPUs
• learning rate and learning rate schedule
• What does the right images augmentation mean?
• training images size

Thank you very much in advance!

Best, Antonin.