Skip to content

Conversation

@LennartPurucker
Copy link
Collaborator

Add support for the TabStar model (https://arxiv.org/abs/2505.18125).

Working on some more TODOs before starting the benchmark

@LennartPurucker
Copy link
Collaborator Author

Related PR for the sake of inversing label cleaner: autogluon/autogluon#5482

@LennartPurucker
Copy link
Collaborator Author

We ran a few benchmarks and are in contact with the first author of the model to improve the method and fix some bugs on their side in the package.

We are waiting for a final method to benchmark and all bugs to be fixed before running the full benchmark.

@LennartPurucker
Copy link
Collaborator Author

Notes for later:

  1. Run TabSTAR wit hthe newest version again
  2. Check if label column name is the correct value or just __label__

# Conflicts:
#	tabarena/pyproject.toml
#	tabflow_slurm/run_setup_slurm_jobs.py
@LennartPurucker
Copy link
Collaborator Author

I updated to the newest version and verified that the label column name is passed to the model.

Once the compute is available, I will rerun its results with tuning and submit this to the official leaderboard, and then merge TabSTAR.

@LennartPurucker
Copy link
Collaborator Author

Started TabArena-Lites runs for a sanity check, and then will go over to full and get results for the submission.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant