search

PJLab-ADG / DetZero

[ICCV 2023] DetZero: Rethinking Offboard 3D Object Detection with Long-term Sequential Point Clouds
https://superkoma.github.io/detzero-page/
Apache License 2.0
317 stars 32 forks source link

ValueError: num_samples should be a positive integer value, but got num_samples=0 #51

Open reynerliu opened 5 months ago

reynerliu commented 5 months ago

When I use the module refining to train GRM: <cd DetZero/refining/tools && python train.py --cfg_file cfgs/ref_model_cfgs/vehicle_grm_model.yaml>,the error broke.I think maybe it is caused by the wrong root of dataset.But I don't know which is the dataset,I check the DATA_CONFIG and I found that: INFO cfg.DATA_CONFIG.DATASET: WaymoGeometryDataset,where is it? Before the above step,I use the command <cd DetZero/daemon && python prepare_object_data.py --track_data_path --split >to prepare object data.As the result,I got the 'refining' direction with three filefolder <'Vehicle','Cyclist','Pedestrian'>,and each of them has several .pkl files. I don't know whether the data which is used for the step to train GRM is generated by the prepare-object-data step?