Hi authors,
Trying to reproduce the LIBERO results. The README references
'LIBERO_data_modified/' but I couldn't find the script that converts
the official LIBERO dataset into this format.
Could you share the preprocessing pipeline, or describe what
modifications were applied (e.g., resolution, camera selection, action
chunking)?
Also, will fine-tuned LIBERO checkpoints be released for verification?
Thanks!
Hi authors,
Trying to reproduce the LIBERO results. The README references
'LIBERO_data_modified/' but I couldn't find the script that converts
the official LIBERO dataset into this format.
Could you share the preprocessing pipeline, or describe what
modifications were applied (e.g., resolution, camera selection, action
chunking)?
Also, will fine-tuned LIBERO checkpoints be released for verification?
Thanks!