could you please provide your training log?

dvlab-research / UVTR

Unifying Voxel-based Representation with Transformer for 3D Object Detection (NeurIPS 2022)

227 stars 18 forks source link

could you please provide your training log? #5

Closed yy9996 closed 2 years ago

yy9996 commented 2 years ago

rt,thanks.

yanwei-li commented 2 years ago

Hi, thanks for your interest. Could you tell me which experiment log you need? Because the previous used workspace is deleted, and only the trained models are preserved. But I'll train the specific model you need again.

yy9996 commented 2 years ago

uvtr_l2c_r101_h11.py ，very appreciated.

yanwei-li commented 2 years ago

No problem. I'll update the results here once finished.

yanwei-li commented 2 years ago

Hi, sorry for the late reply. Here is the running log for UVTR-L2C-R101-H11, because the PyTorch and the mmdet3D version is updated to 1.11 and v1.0, the performance may have some variance. Hope it could help you.

yy9996 commented 2 years ago

谢谢兄弟。