multabench/extended-text-reg-ramen-ratings

REG_TEXT_RAMEN_RATINGS

Part of MulTaBench, a benchmark for multimodal tabular learning.

Paper: MulTaBench: Benchmarking Multimodal Tabular Learning with Text and Image · Leaderboard: multabench/leaderboard

• Target: Stars
• Rows: 4105
• Original source: https://www.kaggle.com/datasets/ankanhore545/top-ramen-ratings-2022

Load the benchmark splits with multabench.load_split("REG_TEXT_RAMEN_RATINGS", fold=0).

Column	Type
Brand	string
Variety	string
Style	categorical
Country	categorical
Stars	float

Citation

@misc{arazi2026multabenchbenchmarkingmultimodaltabular,
  title={MulTaBench: Benchmarking Multimodal Tabular Learning with Text and Image},
  author={Alan Arazi and Eilam Shapira and Shoham Grunblat and Mor Ventura and Elad Hoffer and Gioia Blayer and David Holzmüller and Lennart Purucker and Gaël Varoquaux and Frank Hutter and Roi Reichart},
  year={2026},
  eprint={2605.10616},
  archivePrefix={arXiv},
  primaryClass={cs.LG},
  url={https://arxiv.org/abs/2605.10616},
}