TIGER-Lab/MMEB-train

Massive Multimodal Embedding Benchmark The training data split used for training VLM2Vec models in the paper VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks (ICLR 2025). MMEB benchmark covers 4 meta tasks and 36 datasets meticulously selected for evaluating capabilities of multimodal embedding models. During training, we utilize 20 out of the 36 datasets. For evaluation, we assess performance on the 20 in-domain (IND) datasets and the remaining 16… See the full description on the dataset page: https://huggingface.co/datasets/TIGER-Lab/MMEB-train.

<iframe src="https://silytics.si/embed/item/68db746e-4825-42b2-8335-9d7061b46446" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.