The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
language:
ko license: mit task_categories: text-generation size_categories: 1K<n<10K
korean-pretraining-corpus 한국어 사전학습 코퍼스 Dataset Description 이 데이터셋은 한국어 Foundation Model 학습에 사용된 전처리된 코퍼스입니다. Dataset Summary
Language: Korean (한국어) License: MIT Source: Korean Wikipedia and public Korean text Preprocessing: Tokenized and chunked into sequences
Data Fields
input_ids: 토큰화된 정수 시퀀스 (List[int])
Data Splits
Train: ~95% Validation: ~5%
Dataset Creation Source Data 한국어 위키피디아 및 공개 한국어 텍스트 Preprocessing
텍스트 정제 (최소 길이 필터링) BPE 토큰화 고정 길이 시퀀스로 분할 (512 tokens) 패딩 적용
Tokenization
Tokenizer: BPE (Byte-Pair Encoding)
Vocabulary Size: 32,000
Special Tokens: , , , ,
Usage Loading the Dataset pythonfrom datasets import load_dataset import pickle
HuggingFace에서 다운로드
dataset = load_dataset("oz1115/korean-pretraining-corpus")
또는 로컬 pickle 파일 사용
with open("train.pkl", "rb") as f: train_data = pickle.load(f) Example python# 첫 번째 샘플 sample = dataset['train'][0] print(sample['input_ids'][:10]) # 첫 10개 토큰 Related Models 이 데이터셋으로 학습된 모델:
korean-gpt-150m
Citation bibtex@misc{korean-pretraining-corpus, author = {oz1115}, title = {korean-pretraining-corpus: Korean Pretraining Corpus}, year = {2025}, publisher = {HuggingFace}, url = {https://huggingface.co/datasets/oz1115/korean-pretraining-corpus} } Contact
HuggingFace: @oz1115
- Downloads last month
- 17