---
license: mit
task_categories:
- image-text-to-text
dataset_info:
  features:
  - name: id
    dtype: string
  - name: label
    sequence: string
  - name: image
    sequence: string
  - name: conversations
    list:
    - name: role
      dtype: string
    - name: content
      list:
      - name: type
        dtype: string
      - name: text
        dtype: string
  - name: image_bytes
    sequence: binary
  splits:
  - name: train
    num_bytes: 107899453840
    num_examples: 3776475
  - name: valid
    num_bytes: 28829173
    num_examples: 1000
  download_size: 106379896948
  dataset_size: 107928283013
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
---

This repository contains the dataset used for [JARVIS-VLA: Post-Training Large-Scale Vision Language Models to Play Visual Games with Keyboards and Mouse](https://huggingface.co/papers/2503.16365).

[Project Website](https://craftjarvis.github.io/JarvisVLA/)