llava-next
LLaVA-NeXT-Data
Dataset Card for LLaVA-NeXT
We provide the whole details of LLaVA-NeXT Dataset. In this dataset, we include the data that was used in the instruction tuning stage for LLaVA-NeXT and LLaVA-NeXT(stronger).
Aug 30, 2024: We update the dataset with raw format (de-compress it for json file and images with structured folder), you can directly download them if you are familiar with LLaVA data format.
Dataset Sources
Compared to the instruction data mixture for LLaVA-1.5… See the full description on the dataset page: https://huggingface.co/datasets/lmms-lab/LLaVA-NeXT-Data.LLaVA-NeXT-Interleave-Bench
LLaVA-Interleave Bench Dataset Card
Dataset details
Dataset type:
LLaVA-Interleave Bench is a comprehensive set of multi-image datasets that are collected from public datasets or generated by the GPT-4V API.
It is constructed for evaluating the interleaved multi-image reaoning capbilities of LMMs.
Dataset date:
LLaVA-Interleave Bench was collected in April 2024, and released in June 2024.
Paper or resources for more information:
Blog:… See the full description on the dataset page: https://huggingface.co/datasets/lmms-lab-encoder/LLaVA-NeXT-Interleave-Bench.pred_llava_next_10kLLaVA-NeXT-780k-webdatasetStage-2: Visual Instruction Tuning, this is a subset for quick start.
midjourney-niji-1m-llavanext
Dataset Card for midjourney-niji-1m-llavanext
Dataset Summary
This is a dataset of 2,079,886 synthetic captions for 1,039,943 images from midjourney-v6-520k-raw and nijijourney-v6-520k-raw. The captions were produced using https://huggingface.co/lmms-lab/llama3-llava-next-8b inferenced in float16 after tags were generated with wd-swinv2-tagger-v3, followed by cleanup and shortening with Meta-Llama-3-8B.
All images with metadata are available as MozJPEG encoded JPEGs… See the full description on the dataset page: https://huggingface.co/datasets/CaptionEmporium/midjourney-niji-1m-llavanext.coyo-hd-11m-llavanext
Dataset Card for coyo-hd-11m-llavanext
Dataset Summary
This is a data of 22,794,288 synthetic captions for 11,397,144 images from coyo-700m. The "hd" in the title refers to two aspects: high density and high definition. While large alt-text image pair datasets have many images, only a very small proportion of these images are in higher resolutions and have substantial concept density. For example, many of these datasets consist of more than 50% thumbnail sized or very… See the full description on the dataset page: https://huggingface.co/datasets/CaptionEmporium/coyo-hd-11m-llavanext.