license: unknown language: - ru size_categories: - 100M<n<1B
A synthetic dataset for text recognition tasks, contains 300,000 images.