README
LLaVA-1.5-665K-GenQA-Random-500K
This dataset is used in the paper Zero-Shot Vision Encoder Grafting via LLM Surrogates.
The code for the paper can be found at: https://github.com/kaiyuyue/zero
Dataset Composition
This repository contains the llava-665k_genqa-500k_shuffled.json file, which mixes samples from two sources:
- llava-1.5-665k-instructions without the text-only samples of ShareGPT-40K.
- 500K randomly sampled sequences from genqa.
The order of samples has been already balanced and shuffled.
For more details on the dataset creation and its usage, please refer to the paper and the GitHub repository.
License: Creative Commons Attribution-NonCommercial 4.0 International (CC-BY-NC 4.0)
Similar repos
README
LLaVA-1.5-665K-GenQA-Random-500K
This dataset is used in the paper Zero-Shot Vision Encoder Grafting via LLM Surrogates.
The code for the paper can be found at: https://github.com/kaiyuyue/zero
Dataset Composition
This repository contains the llava-665k_genqa-500k_shuffled.json file, which mixes samples from two sources:
- llava-1.5-665k-instructions without the text-only samples of ShareGPT-40K.
- 500K randomly sampled sequences from genqa.
The order of samples has been already balanced and shuffled.
For more details on the dataset creation and its usage, please refer to the paper and the GitHub repository.
License: Creative Commons Attribution-NonCommercial 4.0 International (CC-BY-NC 4.0)