---
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: tokens
    sequence: string
  - name: pos_tags
    sequence: int64
  splits:
  - name: train
    num_bytes: 2041229.7690445448
    num_examples: 8674
  - name: test
    num_bytes: 874947.2309554551
    num_examples: 3718
  download_size: 873100
  dataset_size: 2916177
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
language:
- udm
task_categories:
- token-classification
size_categories:
- 10K<n<100K
---

The data is taken from the [LingvoDoc](https://lingvodoc.ispras.ru/) website.

['N', 'V', 'PRO', 'CNJ', 'PART', 'NUM', 'IMIT', 'ADV', 'O', 'INTRJ', 'ADJPRO', 'ADJ', 'PREDIC', 'POST', 'ADVPRO', 'PREP', 'PARENTH'] == [0, 1, ..., 16]