Loading...
Loading...
Image-text pairs, instruction tuning, visual QA, cross-modal data, foundation model training data
1,956 datasets
Kaggle hosts the LLaVA model, a multimodal AI system. The dataset likely contains model weights for a large language model with vision capabilities. The author, organization, and last update date are unknown.
The VQA-Autopilot dataset is hosted on Kaggle. Its title suggests it contains data for visual question answering, a task combining computer vision and natural language processing, potentially for applications in autonomous systems. Metadata is minimal; actual content, scale, and authorship require verification after download.
A dataset named LLaVA, hosted on Kaggle, likely contains multimodal data for training vision-language models. The platform tags suggest it is intended for large language model (LLM) training and multimodal AI tasks. Specific details on size, structure, and creation are not provided in the available metadata.
A dataset titled 'data_vlm_diff_ready_30' is hosted on Kaggle. The title suggests it is prepared for training or evaluating vision-language models, likely containing paired image and text data. Its specific content, size, and creation details are not provided in the available metadata.
Adversarial test cases combine images and text to validate multimodal large language models. The dataset is designed to challenge evidence-based reasoning capabilities in models like Gemini. Its origin, size, and creation details are not specified.
A multimodal dataset likely containing transcripts and potentially audio recordings from corporate earnings conference calls. The dataset is hosted on Kaggle, but specific details about its size, source, and creation date are not provided in the available metadata. Its content suggests it is intended for analyzing corporate financial performance and communication.
WBC-AttrDescVQA is a dataset for Visual Question Answering (VQA) tasks, likely involving images of white blood cells (WBCs). The dataset is hosted on Kaggle, but its specific scale, creation date, and authorship are not detailed in the provided metadata. Its content and structure must be verified after download.
A dataset titled 'WildFire_VQA' is hosted on Kaggle. The dataset likely contains image and text pairs for visual question answering tasks related to wildfire scenes. Metadata is minimal; the specific number of samples, data source, and creation date are unknown.
Image captions for nova 2 is a dataset published on Kaggle. The title suggests it likely contains descriptive text paired with images. Metadata is minimal; actual content requires verification after download.
Chart2Code is a benchmark of 2,023 tasks designed to evaluate Large Multimodal Models (LMMs) on chart understanding and code generation, released by CSU-JPG in 2026. The dataset is structured into three hierarchical difficulty levels containing 863, 1,010, and 150 tasks respectively. It maps visual data visualizations to executable code to test the reasoning capabilities of multimodal systems.
RL GSPO Qwen2.5VLM PhaseB Best Composite 180 is a dataset published on Kaggle. The title suggests it is likely a benchmark or evaluation dataset for a vision-language model, possibly related to reinforcement learning. The dataset's specific content, size, and origin are unknown from the provided metadata.
RL GSPO Qwen2.5VLM Staged Code V2 is a dataset hosted on Kaggle. The title suggests it relates to reinforcement learning (RL) and staged training for a vision-language model (VLM) named Qwen2.5. The dataset likely contains data used for training or evaluating such models.
An academic dataset from KAIST, William and Mary, University of Alberta, and Auburn University, released in December 2025. It demonstrates a performance gap in state-of-the-art Vision Language Models (VLMs), which perform perfectly on counting tasks with original images but fail catastrophically on modified versions. The dataset is hosted on Hugging Face by author anvo25.
Orient Anything V2 is an enhanced foundation model for unified understanding of object 3D orientation and rotation from single or paired images. This repository contains the final rendering data used for training the model, as provided by author Viglong. The dataset was last updated on January 13, 2026.
VQA_VIDEOS is a dataset hosted on Kaggle. The title suggests it contains video content paired with questions and answers for visual question answering tasks. The dataset's specific size, content details, and origin are not provided in the available metadata.
Multimodal Emotion Dialogue Dataset is a collection of records for analyzing emotional states in conversations. The dataset likely contains speech, image, and interaction data, as indicated by its raw description. It is hosted on Kaggle, but specific details on its size, creation, and update history are not provided.
Kaggle hosts this dataset titled 'BlipCaptioningOutput'. The title suggests it contains outputs from the BLIP (Bootstrapping Language-Image Pre-training) model, likely pairing images with generated or ground-truth captions. No further metadata on size, source, or creation date is provided.
SciCap Dataset provides pairs of scientific images with corresponding captions. It is designed for training and evaluating multimodal models. The dataset was created for research in scientific image understanding.
A dataset titled 'wavlm_region_model' is hosted on Kaggle. The dataset likely contains audio feature representations or model outputs from the WavLM architecture. Metadata is minimal; actual content, size, and structure require verification after download.
A Kaggle-hosted dataset titled 'wavlm_gender_model'. The dataset's content likely relates to audio data processed by the WavLM architecture for gender classification tasks. Metadata is minimal; the specific number of samples, audio characteristics, and creation details require verification after download.