Question about omitting multi-image data

[prakcoin]

Hi,

I am thinking about using this data for finetuning a VLM to perform autonomous driving. However, I don't think I would need any images other than the front camera, due to system requirement restraints as well as my use case. Would it be fine for me to omit the back-left, back-right, and BEV images, or are there any questions/answers that rely on them?

[IrohXu]

I think you can skip them.

We are working on the v2 version of the dataset and annotations. It will be released here: https://huggingface.co/datasets/LLVM-AD/maplm_v2