LMDrive icon indicating copy to clipboard operation
LMDrive copied to clipboard

[CVPR 2024] LMDrive: Closed-Loop End-to-End Driving with Large Language Models

Results 48 LMDrive issues
Sort by recently updated
recently updated
newest added

Hi, Thanks for open-sourcing your great work. I want to create some custom routes and I saw that in the data generation process, you requires a bunch of .xml configs...

请问有试过放开llm的参数进行训练吗,有试过的话请问跟目前 frozen llm的比精度怎么样?

Hi authors! Thanks for your excellent work. I encounter low training efficiency issues. The second Instruction finetuning stage takes about 6 days on my 8*A100 (40G) GPUs, utilizing only Town01...

Hi, In #21 you give some pointers to the BEV map coordinate, I'm wondering how we can visualize the BEV map processing results made by the vision encoder (like Fig...

生成bev map图时,x,y方向的坐标系是什么? 或者说 measurements里的GPS_x,GPS_y的坐标系是什么? 谢谢

Hi @deepcs233 Thanks for your great work. Our team is investigating this topic and hope to re-implement the codes to **_generate the metric index as your paper claimed_**. However, we...

Hi, I'm trying to run the evaluation through `run_evaluation.sh` on a RTX 3070Ti GPU but the GPU ran out of memory during loading the models, can you share your evaluation...

The specified python version as suggested is 3.8, but after executing the following ``` conda create -n lmdrive python=3.8 conda activate lmdrive cd vision_encoder python setup.py develop ``` I got...

输入的数据的坐标系是怎么定义的? 1. 激光雷达坐标系的定义? ![image](https://github.com/opendilab/LMDrive/assets/19401583/a23f0a63-95da-4ce9-ade6-1bda94c760de) 上图为我保存的输入到模型的激光点云数据,从点云数据来看,激光雷达的坐标系应该是:x指前,y指左,z指上 2. target_point坐标系定义? ![image](https://github.com/opendilab/LMDrive/assets/19401583/22aa574e-6acd-45be-a23a-73ddda769aa3) 上图为我保存的输入到模型的target_point的打印,此时车辆在右转,根据打印坐标看,我理解的target_point的坐标系定义是:X指右,y指后,因为车辆在直道上行驶的时候x几乎为零,y是从-50到0递增。 3. 还有计算target_point的时候用到的theta是怎么定义的?和是和那个轴的夹角,逆时针还是顺时针为正? ![image](https://github.com/opendilab/LMDrive/assets/19401583/24a52b52-b899-4a70-a555-ba27d59855f2) 4. 车身坐标系的定义和target_point一样吗? 以上坐标系的定义看起来和carla里面不一样,我这边查到的carla里面的坐标系定义如下图: ![image](https://github.com/opendilab/LMDrive/assets/19401583/7ee03cf1-8a50-40b8-b227-a685c95dadd5) 我目前对于代码里坐标系定义这块还不是很清楚,能否明确一下?