CoolFace
Datasetpublic

Polygl0t/mc4-50k

mC4 (50K samples per language) Dataset Summary This is the multilingual version of the C4 dataset. We only keep 50,000 samples per language for all the 108 languages available. This is intended for educational and experimentation purposes. For the full dataset, please refer to the multilingual config of the allenai/c4 dataset. How do I download this? Using ๐Ÿค— Datasets from datasets import load_dataset # Portuguese only pt =โ€ฆ See the full description on the dataset page: https://huggingface.co/datasets/Polygl0t/mc4-50k.

sourceHugging Faceupdated 9mo agoView on Hugging Face
0likes72downloads
Dataset Card

mC4 (50K samples per language)

Dataset Description

  • โ€”Paper: https://arxiv.org/abs/1910.10683

Dataset Summary

This is the multilingual version of the C4 dataset. We only keep 50,000 samples per language for all the 108 languages available. This is intended for educational and experimentation purposes. For the full dataset, please refer to the `multilingual` config of the `allenai/c4` dataset.

How do I download this?
Using ๐Ÿค— Datasets
python
from datasets import load_dataset

# Portuguese only
pt = load_dataset("Polygl0t/mc4-50k", "pt")

# All languages (108)
mc4 = load_dataset("Polygl0t/mc4-50k")

You can also load and mix multiple languages:

python
from datasets import concatenate_datasets, interleave_datasets, load_dataset

es = load_dataset("Polygl0t/mc4-50k", "es", streaming=True)
fr = load_dataset("Polygl0t/mc4-50k", "fr", streaming=True)

# Concatenate both datasets
concatenated = concatenate_datasets([es, fr])
# Or interleave them (alternates between one and the other)
interleaved = interleave_datasets([es, fr])
Languages

Note that the languages that end with "-Latn" are simply romanized variants, i.e. written using the Latin script.

language codelanguage name
afAfrikaans
amAmharic
arArabic
azAzerbaijani
beBelarusian
bgBulgarian
bg-LatnBulgarian (Latin)
bnBangla
caCatalan
cebCebuano
coCorsican
csCzech
cyWelsh
daDanish
deGerman
elGreek
el-LatnGreek (Latin)
enEnglish
eoEsperanto
esSpanish
etEstonian
euBasque
faPersian
fiFinnish
filFilipino
frFrench
fyWestern Frisian
gaIrish
gdScottish Gaelic
glGalician
guGujarati
haHausa
hawHawaiian
hiHindi
hi-LatnHindi (Latin script)
hmnHmong, Mong
htHaitian
huHungarian
hyArmenian
idIndonesian
igIgbo
isIcelandic
itItalian
iwformer Hebrew
jaJapanese
ja-LatnJapanese (Latin)
jvJavanese
kaGeorgian
kkKazakh
kmKhmer
knKannada
koKorean
kuKurdish
kyKyrgyz
laLatin
lbLuxembourgish
loLao
ltLithuanian
lvLatvian
mgMalagasy
miMaori
mkMacedonian
mlMalayalam
mnMongolian
mrMarathi
msMalay
mtMaltese
myBurmese
neNepali
nlDutch
noNorwegian
nyNyanja
paPunjabi
plPolish
psPashto
ptPortuguese
roRomanian
ruRussian
ru-LatnRussian (Latin)
sdSindhi
siSinhala
skSlovak
slSlovenian
smSamoan
snShona
soSomali
sqAlbanian
srSerbian
stSouthern Sotho
suSundanese
svSwedish
swSwahili
taTamil
teTelugu
tgTajik
thThai
trTurkish
ukUkrainian
undUnknown language
urUrdu
uzUzbek
viVietnamese
xhXhosa
yiYiddish
yoYoruba
zhChinese
zh-LatnChinese (Latin)
zuZulu

Dataset Structure

Data Instances

An example form the en config is:

{
  'text': 'Beginners BBQ Class Taking Place in Missoula!\nDo you want to get better at making delicious BBQ? You will have the opportunity, put this on your calendar now. Thursday, September 22nd join World Class BBQ Champion, Tony Balay from Lonestar Smoke Rangers. He will be teaching a beginner level class for everyone who wants to get better with their culinary skills.\nHe will teach you everything you need to know to compete in a KCBS BBQ competition, including techniques, recipes, timelines, meat selection and trimming, plus smoker and fire information.\nThe cost to be in the class is $35 per person, and for spectators it is free. Included in the cost will be either a t-shirt or apron and you will be tasting samples of each meat that is prepared.'
}

Data Fields

The data have several fields:

  • โ€”text: text content as a string

Source Data

Initial Data Collection and Normalization

The C4 and mC4 datasets are collections text sourced from the public Common Crawl web scrape. It includes heuristics to extract only natural language (as opposed to boilerplate and other gibberish) in addition to extensive deduplication.

Licensing Information

We are releasing this dataset under the terms of ODC-BY. By using this, you are also bound by the Common Crawl terms of use in respect of the content contained in the dataset.

Polygl0t/mc4-50k ยท CoolFace