| license: mit | |
| task_categories: | |
| - image-to-text | |
| language: | |
| - ru | |
| size_categories: | |
| - 100K<n<1M | |
| Dataset generated from HKR train set using Stackmix | |
| =================================================== | |
| Number of images: 300000 | |
| Sources: | |
| * [HKR dataset](https://github.com/abdoelsayed2016/HKR_Dataset) | |
| * [Stackmix code](https://github.com/ai-forever/StackMix-OCR) | |