---
license: mit
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
dataset_info:
  features:
  - name: text
    dtype: string
  - name: label
    dtype:
      class_label:
        names:
          '0': negative
          '1': positive
  splits:
  - name: train
    num_bytes: 65009575.70216127
    num_examples: 278021
  - name: test
    num_bytes: 16252569.297838729
    num_examples: 69506
  download_size: 45469780
  dataset_size: 81262145
task_categories:
- text-classification
language:
- ru
tags:
- russian
- toxic
- classification
size_categories:
- 100K<n<1M
---

# Russian toxic text datasets
This datasets is a merge of
- [AlexSham/Toxic_Russian_Comments](https://huggingface.co/datasets/AlexSham/Toxic_Russian_Comments)
- [marriamaslova/toxic_dvach](https://huggingface.co/datasets/marriamaslova/toxic_dvach)
- [textdetox/multilingual_toxicity_dataset](https://huggingface.co/datasets/textdetox/multilingual_toxicity_dataset)
- Parsed toxic and non toxic texts from VK Stream API

There is more than 300k toxic and non toxic comments