Datasets:
metadata
dataset_info:
features:
- name: image
dtype: image
- name: conversations
dtype: string
splits:
- name: train
num_bytes: 46019106088.205
num_examples: 272795
download_size: 20289135489
dataset_size: 46019106088.205
task_categories:
- visual-question-answering
language:
- en
tags:
- multimodal
- vision
size_categories:
- 100K<n<1M
license: cc-by-nc-4.0
LLaVA Instruct Mix
Added OCR and Chart QA dataset into this for more text extraction questions