Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
sozkz-corpus-clean-enkk-fineweb-edu-v1 – Dataset by stukenov | AlphaNeural AI
You can deploy this model and start earning money today!
stukenov
/
sozkz-corpus-clean-enkk-fineweb-edu-v1
like
0
translation
HuggingFaceFW/fineweb-edu-score-2
kk
en
cc-by-4.0
10M<n<100M
parquet
optimized-parquet
text
datasets
dask
polars
mlcroissant
us
kazakh
english
machine-translation
fineweb-edu
parallel-corpus
sozkz
Views
No views yet
Model card
Files and Versions
Community
API
SozKZ Corpus Clean EN→KK (FineWeb-Edu) v1
Machine-translated parallel corpus of educational web texts (English → Kazakh), derived from FineWeb-Edu.
Overview
Source FineWeb-Edu Score-2 (first 20M rows)
Languages English (en) → Kazakh (kk)
Translation CTranslate2, NLLB-200 (en→kk), greedy decoding
Filtering Length (50–10K chars), exact dedup (xxhash), language detection (fasttext)
Rows ~18M (after filtering ~90% pass rate)
Format Expandable… See the full description on the dataset page:
https://huggingface.co/datasets/stukenov/sozkz-corpus-clean-enkk-fineweb-edu-v1
.
