Skip to content

3D Object Detection using BEVFormer (tiny)

Dataset

The benchmark implementation run command will automatically download the validation and calibration datasets and do the necessary preprocessing. In case you want to download only the datasets, you can use the below commands.

BEVFormer validation run uses the NuScenes dataset. The preprocessed data is located in nuscenes_data/preprocessed/val_3d.tar.gz.

Get Validation Dataset

mlcr get,preprocessed,dataset,nuscenes,_mlc,_validation --outdirname=<path_to_download>

Get Calibration Dataset

mlcr get,preprocessed,dataset,nuscenes,_mlc,_calibration --outdirname=<path_to_download>

For preprocessing the dataset yourself, you can download the raw dataset.

Get Raw Dataset

mlcr get,dataset,nuscenes,_mlc,_rclone --outdirname=<path_to_download>

Model

The benchmark implementation run command will automatically download the required model and do the necessary conversions. In case you want to only download the official model, you can use the below commands.

Get the Official MLPerf BEVFormer Model

ONNX

mlcr get,ml-model,bevformer,_mlc,_rclone,_onnx --outdirname=<path_to_download>

PyTorch

mlcr get,ml-model,bevformer,_mlc,_rclone,_pytorch --outdirname=<path_to_download>