README.md

metadata

license: mit
task_categories:
  - image-to-text
language:
  - ru
size_categories:
  - 100K<n<1M

Dataset generated from cyrillic train set using Stackmix

Number of images: 300000

Sources:

• Cyrillic dataset
• Stackmix code