No such file or directory: 'data/KITTI/samples.pkl'

[HaiCLi]

Hi,

I used the following commands to download the dataset and make the folder tree like this. However, I didn't find the 'data/KITTI/samples.pkl'. Is there anything I missed?

commands for downloading: "# Velodyne data wget https://s3.eu-central-1.amazonaws.com/avg-kitti/data_object_velodyne.zip

Calib mats

wget https://s3.eu-central-1.amazonaws.com/avg-kitti/data_object_calib.zip

Training labels

wget https://s3.eu-central-1.amazonaws.com/avg-kitti/data_object_label_2.zip

Dev kit for reference

wget https://s3.eu-central-1.amazonaws.com/avg-kitti/devkit_object.zip

Sequence data

wget https://s3.eu-central-1.amazonaws.com/avg-kitti/devkit_road_mapping.zip

Download splits

wget -c https://raw.githubusercontent.com/traveller59/second.pytorch/master/second/data/ImageSets/test.txt wget -c https://raw.githubusercontent.com/traveller59/second.pytorch/master/second/data/ImageSets/train.txt wget -c https://raw.githubusercontent.com/traveller59/second.pytorch/master/second/data/ImageSets/val.txt wget -c https://raw.githubusercontent.com/traveller59/second.pytorch/master/second/data/ImageSets/trainval.txt

then unzip everything"

Folder tree:

Thanks for your help in advance!

[willGuimont]

Hi,

Thanks for raising this issue! I’ve updated the documentation in docs/TRAINING.md to clarify how to generate data/KITTI/samples.pkl.

To generate the file, you can run the following command:

PYTHONPATH=. python scripts/generate_kitti_object_sampler.py

Let me know if you have any other questions!

[HaiCLi]

Hi,

Thanks for raising this issue! I’ve updated the documentation in docs/TRAINING.md to clarify how to generate data/KITTI/samples.pkl.

To generate the file, you can run the following command:

PYTHONPATH=. python scripts/generate_kitti_object_sampler.py

Let me know if you have any other questions!

I think this command can not solve this problem. It seems that running this command requires "data/KITTI/samples.pkl", rather than outputting the file. "data/KITTI/samples.pkl" is the prerequirement.

[willGuimont]

Hi,

I think the issue stems from the fact that data/KITTI/ does not exist, hence the No such file or directory error.

Are you running the code in Docker? Did you add the volume containing the data to the correct location? It can be added using the -v ./data/KITTI:data/KITTI flag in Docker.

From the root of the project, you should be able to list the content of KITTI:

ls data/KITTI

Otherwise, I would suggest following the updated instructions in docs/TRAINING.md, with the most up-to-date version of the code:

docker build -t mask_bev .
export CUDA_VISIBLE_DEVICES=0

docker run --gpus all -e CUDA_VISIBLE_DEVICES=$CUDA_VISIBLE_DEVICES --rm -it \
  -v .:/app \
  -v ./data/KITTI:/app/data/KITTI \
  -v /dev/shm:/dev/shm \
  mask_bev bash -c "PYTHONPATH=. python3.10 scripts/generate_kitti_object_sampler.py"

Let me know if it works!

[HaiCLi]

Hi,

I think the issue stems from the fact that data/KITTI/ does not exist, hence the No such file or directory error.

Are you running the code in Docker? Did you add the volume containing the data to the correct location? It can be added using the -v ./data/KITTI:data/KITTI flag in Docker.

From the root of the project, you should be able to list the content of KITTI:

ls data/KITTI

Otherwise, I would suggest following the updated instructions in docs/TRAINING.md, with the most up-to-date version of the code:

docker build -t mask_bev .
export CUDA_VISIBLE_DEVICES=0

docker run --gpus all -e CUDA_VISIBLE_DEVICES=$CUDA_VISIBLE_DEVICES --rm -it \
  -v .:/app \
  -v ./data/KITTI:/app/data/KITTI \
  -v /dev/shm:/dev/shm \
  mask_bev bash -c "PYTHONPATH=. python3.10 scripts/generate_kitti_object_sampler.py"

Let me know if it works!

I solved this by reinstall the requirement.txt. But didn't know why, whatever it was solved.