OpenOrca-KO

When this dataset fits

OpenOrca-KO is a Korean Instruction Tuning Dataset from Hugging Face. It fits instruction tuning, not long-document pretraining. License label MIT still needs review. Size text is “21.8MB”, copied from the host.

Task / category

Text · Text · LLM, reasoning

Language

Korean

Format · Size

Text · 21.8MB (21,632)

License

MIT(Catalog hint — re-check original) · License document

Source & original link

Hugging Face · Original dataset link(Verified)

Verification status

Source: Verified · License: Needs review · Last checked: 2026-09-03

OpenOrca-KO: Korean Instruction Tuning Dataset from Hugging Face. License MIT. Open the original link to download.

OpenOrca-KO (2023). Hugging Face. license label: MIT catalog commercial mark: allowed Original: https://huggingface.co/datasets/kyujinpy/OpenOrca-KO. GearDel catalog: https://geardel.com/en/datasets/openorca-ko. Last checked: 2026-09-03. Author names are not stored in the GearDel catalog and are not invented here.

Differences below use catalog fields only: license, commercial mark, provider, size text, and year.

• HAE-RAE Bench 2.0Korean LLM Benchmark Dataset · Hugging FaceViewSize text differs (21.8MB vs 1.0MB). Those strings are copied from the hosts. Catalog years differ (2023 vs 2024).
• HAE-RAE CoT 1.5MKorean NLP Dataset · Hugging FaceViewLicense labels differ (MIT vs CC BY 4.0). Size text differs (21.8MB vs 1.05GB). Those strings are copied from the hosts. Catalog years differ (2023 vs 2024).
• KIT-19Korean Instruction Tuning Dataset · Hugging FaceViewLicense labels differ (MIT vs CC BY 4.0). Size text differs (21.8MB vs 85MB). Those strings are copied from the hosts. Catalog years differ (2023 vs 2024).
• KoAlpaca v1.0 AlpacaKorean Instruction Tuning Dataset · Hugging FaceViewLicense labels differ (MIT vs CC BY-NC 4.0). Commercial marks differ (Commercial use allowed (catalog) vs Commercial use may be conditional). Size text differs (21.8MB vs 8.1MB). Those strings are copied from the hosts.