---
dataset_info:
  features:
  - name: label
    dtype:
      class_label:
        names:
          '0': negative
          '1': positive
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  splits:
  - name: train
    num_bytes: 11392188496
    num_examples: 278021
  - name: test
    num_bytes: 2848077856
    num_examples: 69506
  download_size: 38238491
  dataset_size: 14240266352
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
license: mit
task_categories:
- text-classification
language:
- ru
tags:
- tokenized
- toxic
- russian
- classify
size_categories:
- 100K<n<1M
---

# Tokenized Russian toxic text
Tokenized version of [Mnwa/russian-toxic](https://huggingface.co/datasets/Mnwa/russian-toxic) dataset with modernbert base model tokenizer