Problems training train_single_seq

pxiangwu / MotionNet

CVPR 2020, "MotionNet: Joint Perception and Motion Prediction for Autonomous Driving Based on Bird's Eye View Maps"

169 stars 25 forks source link

Problems training train_single_seq #4

Closed HenryJunW closed 4 years ago

HenryJunW commented 4 years ago

Impressive work.

When I tried to train the model without using spatio-temporal consistency losses from scratch, I found out that the training dataset cannot be imported successfully. It turned out that there is a typo in the line 171 of MotionNet/data/nuscenes_dataloader.py 'os.path.isfile(os.path.join(self.dataset_root, d))]', which is supposed to be 'os.path.isdir(os.path.join(self.dataset_root, d))]'.

BTW, it would be very helpful if you can provide a pre-trained single-seq model. In addition, how long does it take to train the model?

pxiangwu commented 4 years ago

Hi, yes, that is a typo and should have been fixed in my original repo on Github instead of that on MERL (although my original repo was removed ...). And I also provide pre-trained models:

The pre-trained model for train_multi_seq.py can be downloaded from Google Drive or Dropbox
The pre-trained model for train_multi_seq_MGDA.py can be downloaded from Google Drive or Dropbox

The training usually takes less than one day on a single RTX 2080 Ti GPU.

HenryJunW commented 4 years ago

Thanks for your instant response. Appreciate it.