When will the MLLM-4D-Bench evaluation JSON and code be open-sourced?

[wenzhoulyu]

Hi, thanks for the great work on MLLM-4D!

We noticed the training and evaluation videos are already available on Hugging Face, but the JSON annotations and evaluation code required to run MLLM-4D-Bench don't seem to be public yet.

Could you let us know when you plan to open-source the evaluation JSON files and the eval code? We'd really like to reproduce the benchmark results and build on your work.

Thanks in advance!