---
dataset_info:
  features:
  - name: text
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  splits:
  - name: train
    num_bytes: 5640107996
    num_examples: 2169710
  download_size: 2834538255
  dataset_size: 5640107996
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

# Open Voyage Text to Vector Dataset 

<p align="center">
  <img src="./txt2vec.png" width="512px" height="512px" />
</p>

A ShareGPT Style of around 3 million SVG images with their BLIP2 captioned prompts.