metadata
dataset_info:
features:
- name: messages
large_list:
- name: content
dtype: large_string
- name: role
dtype: large_string
- name: source_dataset
dtype: large_string
splits:
- name: train
num_bytes: 1452460394
num_examples: 515911
download_size: 708508292
dataset_size: 1452460394
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: cc-by-sa-4.0
task_categories:
- question-answering
language:
- ko
source_datasets:
- coastral/korean-writing-style-instruct
- lemon-mint/smol-koreantalk
- CarrotAI/ko-instruction-dataset
- MarkrAI/KOpen-HQ-Hermes-2.5-60K
- beomi/KoAlpaca-RealQA
- HAERAE-HUB/qarv-instruct-ko
- maywell/koVast
devngho/korean-instruction-mix
한국어 instruction 데이터셋 mix. 이 데이터셋을 사용해 학습된 모델은 아래 원본 데이터셋을 함께 인용해주세요.
- coastral/korean-writing-style-instruct(apache-2.0): 5.62%, x1.0
- lemon-mint/smol-koreantalk(apache-2.0): 26.77%, x0.3
- CarrotAI/ko-instruction-dataset(apache-2.0): 2.73%, x2.0
- MarkrAI/KOpen-HQ-Hermes-2.5-60K(mit): 23.28%, x2.0
- beomi/KoAlpaca-RealQA(cc by-sa 4.0): 7.18%, x2.0
- HAERAE-HUB/qarv-instruct-ko(apache-2.0): 7.89%, x4.0
- maywell/koVast(mit): 26.54%, x0.2