datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
kurdish-corpus
Kurdish Corpus
A large-scale multi-source Kurdish language dataset for training language models.
Dataset Statistics
Total Documents: 1,797,686
Total Tokens: 625,716,980
Shards: 4
Built: 2026-05-02
By Language
Language
Documents
Sorani (ckb)
1,274,425
Kurmanji (kmr)
478,540
Zazaki (diq)
34,069
Hawrami (hac)
10,652
By Source Type
Source Type
Documents
News
1,443,750
Web
229,705
Wikipedia
124,231
Usage… See the full description on the dataset page: https://huggingface.co/datasets/kurdish-ai/kurdish-corpus.kurdish-web
Dataset Card for Kurdish Web Corpus (Deduplicated)
Dataset Summary
A multilingual corpus of web text in three Kurdish/Zaza varieties — Kurmanji Kurdish
(kmr_Latn), Sorani Kurdish (ckb_Arab), and Zazaki (diq_Latn) — scraped from Kurdish-language
websites, language-identified with GlotLID, and
deduplicated (exact + MinHash near-duplicate removal). See the "Dataset Creation" section below
for the full pipeline.
Rows, by language config:
config
language
script… See the full description on the dataset page: https://huggingface.co/datasets/muzaffercky/kurdish-web.KurdishCorpus-Clean
⚠️ This is a personal mirror. The actively maintained, canonical version of this dataset is kurdish-tech/KurdishCorpus-clean — that's the one to cite, link, and build on. This copy is kept for history.
The Largest Documented Kurmanji and Multi-Dialect Kurdish Dataset — v1.1
A large-scale, multi-dialect Kurdish text corpus prioritizing native Kurmanji
(Northern Kurdish) fluency, with substantial Sorani (Central Kurdish) and a
Zazaki baseline. Built for language modeling… See the full description on the dataset page: https://huggingface.co/datasets/alanhasn/KurdishCorpus-Clean.kurdish-kurmanji-theses
Kurdish Kurmanji Academic Theses
A corpus of Kurdish Kurmanji (Northern Kurdish, Latin script) text extracted from academic theses indexed in YÖK Tez Merkezi, the Turkish national thesis repository.
Dataset Summary
Each record is a single thesis whose body text has been filtered to Kurmanji-only paragraphs. Non-Kurdish paragraphs (Turkish, Arabic, etc.) and structural noise (figure captions, table of contents entries, bare page numbers) were removed using the… See the full description on the dataset page: https://huggingface.co/datasets/muzaffercky/kurdish-kurmanji-theses.KurdishCorpus-clean
The Largest Documented Kurmanji and Multi-Dialect Kurdish Dataset — v1.1
A large-scale, multi-dialect Kurdish text corpus prioritizing native Kurmanji
(Northern Kurdish) fluency, with substantial Sorani (Central Kurdish) and a
Zazaki baseline. Built for language modeling, tokenizer training, and
general-purpose Kurdish NLP.
This release contains only openly-licensed or presumptively-free
redistributable content. A parallel research-tier subset (copyrighted
commercial… See the full description on the dataset page: https://huggingface.co/datasets/kurdish-tech/KurdishCorpus-clean.central-kurdish-correction-table
Central Kurdish Orthographic Correction Table
This repository contains a correction table designed to standardize Central Kurdish (Sorani) script for speech and language technology applications.
Orthographic variation is a major challenge in Kurdish NLP. Different spellings and writing conventions for the same words can introduce inconsistencies in ASR training, evaluation, and downstream NLP systems.
This table provides mappings from non-standard or inconsistent forms to their… See the full description on the dataset page: https://huggingface.co/datasets/aranemini/central-kurdish-correction-table.aya-processed-central-kurdish-paraphrase
Aya Processed Central Kurdish Paraphrase Identification Dataset
Dataset Description
This dataset contains 49,401 preprocessed Central Kurdish paraphrase identification samples in conversational format, derived from the Aya Collection Language Split.
Languages
Central Kurdish (ckb)
Dataset Structure
The dataset contains the following columns:
id: Unique identifier for each sample
prompt: Conversational format as JSON array containing the… See the full description on the dataset page: https://huggingface.co/datasets/shiima/aya-processed-central-kurdish-paraphrase.Kurdish_Medical_Corpus_KMC
Kurdish Medical Corpus (KMC) V1
The Kurdish Medical Corpus (KMC) V1 is a high-quality, domain-specific dataset designed for instruction-tuning and scientific knowledge discovery in Central Kurdish (Sorani). This dataset consists entirely of human-written medical content, avoiding the common pitfalls of machine-translated corpora in low-resource language research.
Dataset Summary
KMC V1 is concentrated into a single, consolidated JSON file: kurdish_medical_corpus_kmc.json.… See the full description on the dataset page: https://huggingface.co/datasets/shkomq/Kurdish_Medical_Corpus_KMC.Roleplay-Kurdish
RolePlay-Kurdish
Roleplay-Kurdish Dataset is a dataset for roleplaying in the Kurdish language for the Large Language Model.
The base dataset is GPTeacher role play dataset by teknium 1, which can be found under this link, released under MIT License. The dataset is then translated into respective languages. The translation process is powered by Google Translate, using cloud translation API.
For more information and other language datasets for roleplay, see this github repo.
For… See the full description on the dataset page: https://huggingface.co/datasets/jojo-ai-mst/Roleplay-Kurdish.kurdish-sorani-corpuskurdish-ai-dataset
Kurdish Corpus
A large-scale multi-source Kurdish language dataset for training language models.
Dataset Statistics
Total Documents: 1,797,686
Total Tokens: 625,716,980
Shards: 4
Built: 2026-05-02
By Language
Language
Documents
Sorani (ckb)
1,274,425
Kurmanji (kmr)
478,540
Zazaki (diq)
34,069
Hawrami (hac)
10,652
By Source Type
Source Type
Documents
News
1,443,750
Web
229,705
Wikipedia
124,231
Usage… See the full description on the dataset page: https://huggingface.co/datasets/kurdish-ai/kurdish-ai-dataset.kurdish-bidirectional-translation-finall
Kurdish Bidirectional Translation Dataset
Overview
This dataset contains bidirectional translation pairs for Kurdish (Central Kurdish - Sorani script) with English, Arabic, and Persian.
Each sample includes:
translation: Forward translation (Source → Kurdish)
back_translation: Reverse translation (Kurdish → Source)
Dataset Details
Total Samples: 627,103
Languages: Kurdish (ckb), English (en), Arabic (ar), Persian (fa)
Format: JSON conversations with… See the full description on the dataset page: https://huggingface.co/datasets/shiima/kurdish-bidirectional-translation-finall.kurdish-unified-corpus
Unified Kurdish Corpus
Dataset Description
This dataset aggregates 758,166 Kurdish text samples from multiple high-quality sources. All text has been preprocessed using asosoft.
Languages
Central Kurdish (ckb)
Kurdish (ku)
Dataset Structure
Columns
text: Preprocessed text content (asosoft applied)
base_dataset: Source dataset name
url: Source URL (NULL if not available)
word_count: Number of words (space-separated)
character_count:… See the full description on the dataset page: https://huggingface.co/datasets/shiima/kurdish-unified-corpus.Kurdish_Multi-Domain_Corpus_KMDC
Kurdish Multi-Domain Corpus (KMDC)
Dataset Description
The Kurdish Multi-Domain Corpus (KMDC) is a large-scale instruction-style dataset designed to support natural language processing (NLP), supervised fine-tuning (SFT), and large language model (LLM) development for Central Kurdish (Sorani). The dataset consists of structured question–response pairs generated through an LLM-guided pipeline that transforms raw Kurdish text into machine-learning-ready… See the full description on the dataset page: https://huggingface.co/datasets/shkomq/Kurdish_Multi-Domain_Corpus_KMDC.kurdish-train-small