Linyan Huang

Hi, we will soon release a model for planning tasks.

Hello @zc-zhao, our training dataset consists of 9 million pieces of data sourced from the open world, which can refer to the Tab. 2 in the paper. The training configuration...

Hi, thanks for your interest. 1. `BOXQA_train_v3.json` is essentially the same as `box_detection_train.json`, accessible through this [link](https://drive.google.com/drive/folders/1QFBIrKqxjn9lfv31XMC3wVIdaAbpMwDL). 2. In the Traffic Sign Inquiry, it denotes the time needed for recognition....

Hi @pastapeter, thanks for pointing it out. We will fix it.

Thank you for your attention. We expect to release models gradually before the midpoint of the year.

Hi @fang196, the files `bevdetv2-nuscenes_infos_trainval.pkl` have been uploaded to this [link](https://drive.google.com/drive/folders/1QFBIrKqxjn9lfv31XMC3wVIdaAbpMwDL). DriveLM is conducting the competition using the file `data/drivelm/val.json`. We will release it after the competition concludes.

Could you provide more details such as your files and code? I can not reproduce your error now.

@piqiuni If it doesn't work, you may revert to the previous version of the code prior to the pull request [#60](https://github.com/OpenDriveLab/DriveLM/pull/60).

You could employ post-processing techniques, such as rule-based methods, to address this situation. For instance, you could opt to retain only the initial 12 numbers. Subsequently, you could proceed with...

The error you're encountering is actually caused by your own model. Typically, the model should produce the even number as output.