README.md

---
language:
- fa
size_categories:
- 1K<n<10K
task_categories:
- token-classification
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: text
    dtype: string
  - name: relation
    dtype: string
  - name: e1
    dtype: string
  - name: e2
    dtype: string
  - name: is_reversed
    dtype: bool
  - name: raw
    dtype: string
  splits:
  - name: train
    num_bytes: 3306893
    num_examples: 8000
  - name: test
    num_bytes: 1113759
    num_examples: 2717
  download_size: 2331411
  dataset_size: 4420652
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---

Link to article: [https://arxiv.org/abs/2005.06588](https://arxiv.org/abs/2005.06588)

Downloaded data from [farsbase.net](http://farsbase.net/PERLEX.html)