Request: Pre-computed object representations for additional BOP datasets

Hi,

Thank you for releasing FoundPose and the accompanying code. I have been
able to successfully reproduce the reported LMO results using the provided
`object_repre.zip`.

I am currently extending FoundPose as part of a research project and would
like to evaluate on additional BOP datasets (e.g., YCBV, TLESS, HB).
However, since K-means clustering and PCA involve random initialization,
re-running `gen_repre.py` from scratch produces different visual word
vocabularies each time, making it difficult to fairly reproduce your
baseline numbers on those datasets.

Would it be possible to share the pre-computed object representations
(`repre.pth` files) for other BOP datasets, similar to the LMO release?

If sharing via GitHub is not feasible, feel free to send the files to:
**mooread3031@gmail.com**

Thank you for your time and for the great work!