Dataset · Hugging Face

gh0stHunter/MMR-Bench-V2

Name: MMR-Bench-V2
Creator: gh0stHunter
License: other
Keywords: task_categories:visual-question-answering, language:en, language:zh, license:other, size_categories:1M<n<10M, format:parquet, modality:tabular, modality:text, library:datasets, library:dask, library:polars, library:mlcroissant, region:us, multimodal, model-routing

MMR-Bench V2 Dataset Description MMR-Bench V2 is a normalized dataset for offline multimodal model-routing research. It joins canonical multimodal benchmark instances with previously collected model outcomes so that routing methods can learn or evaluate which vision-language model to select for each input. Release v2.0.3 contains: 25,504 instances from 18 benchmarks 1,122,176 model-result records covering 44 models 1,118,161 successful results and 4,015 explicit… See the full description on the dataset page: https://huggingface.co/datasets/gh0stHunter/MMR-Bench-V2.

View on Hugging FaceReport an issue

Author / organization	gh0stHunter
License (as declared)	other
Downloads	1.4K (as reported by Hugging Face)
Likes	1
Created	23 July 2026
Last updated at source	30 September 2026

multimodalmodel-routingbenchmarkvlmevalkitoffline-evaluation

Embed this entry

<iframe src="https://silytics.si/embed/item/72a1f160-0ea1-4260-8791-3b14e7f01c22" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 2 days

1,365 now (+26 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

More from gh0stHunter

gh0stHunter/MMR-BenchDataset

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.