autogluon / autogluon/autogluon
[multimodal] Enable tabular foundational models for AutoMM use cases
Open
enhancement
module: multimodal
module: tabular
- Dominant language
- Python
- Stars
- 10.7k
- Forks
- 1.2k
- Avg merge
- 21h 29m
- Merged PRs (30d)
- 57
Description
## Description
Currently tabular foundational models such as tabicl, mitra and tabpfnv2 are available in autogluon.tabular. It would be interesting to explore and benchmark if they can enhance automm as well
## References
https://github.com/autogluon/autogluon/pull/5227#issuecomment-3089091137
Contributor guide
Research direction
Start by reviewing the existing tabicl, mitra, and tabpfnv2 implementations in autogluon.tabular and the AutoMM use cases. Define a benchmark that compares their effect in AutoMM, then document the results and integration requirements; the linked pull request comment provides the stated reference.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 30/100