| 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](https://www.kaggle.com/datasets/constantinwerner/cyrillic-handwriting-dataset) | |
| * [Stackmix code](https://github.com/ai-forever/StackMix-OCR) | |