Datasets:

blue-machines
/

lid_all_scripts_dataset

license: other
task_categories:
  - text-classification
language:
  - en
  - hi
  - bn
  - gu
  - kn
  - ml
  - mr
  - or
  - ta
  - te
tags:
  - language-identification
  - indic
  - code-mixed
  - romanized
  - native-script
pretty_name: LID All Scripts Dataset
size_categories:
  - 1M<n<10M
dataset_info:
  features:
    - name: id
      dtype: string
    - name: text
      dtype: string
    - name: type
      dtype: string
    - name: language
      dtype: string
  splits:
    - name: train
      num_bytes: 798415525
      num_examples: 2897517
  download_size: 307746472
  dataset_size: 798415525
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

blue-machines/lid_all_scripts_dataset

All-scripts language identification corpus derived from unified_dataset.csv.

Columns

id
text
type
language

Stats

Rows: 2,897,517
Source file: unified_dataset.csv

Load

from datasets import load_dataset

ds = load_dataset("blue-machines/lid_all_scripts_dataset")
print(ds)
print(ds["train"][0])