Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/NomaDamas/awesome-korean-llm
Awesome list of Korean Large Language Models.
https://github.com/NomaDamas/awesome-korean-llm
List: awesome-korean-llm
Last synced: about 1 month ago
JSON representation
Awesome list of Korean Large Language Models.
- Host: GitHub
- URL: https://github.com/NomaDamas/awesome-korean-llm
- Owner: NomaDamas
- License: cc0-1.0
- Created: 2023-08-15T16:25:37.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2023-10-31T10:35:42.000Z (about 1 year ago)
- Last Synced: 2024-05-22T20:03:45.873Z (7 months ago)
- Size: 19.5 KB
- Stars: 402
- Watchers: 10
- Forks: 27
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- ultimate-awesome - awesome-korean-llm - Awesome list of Korean Large Language Models. (Other Lists / Monkey C Lists)
README
# Awesome Korean LLM [](https://awesome.re)
한국어 오픈소스 LLM 정보를 모아놓은 awesome list입니다.
## 한국어 오픈소스 LLM
| 이름 | 사이즈 | 제작자 | 베이스 모델 | 상업적 사용 여부 | 가중치 |
|:----------------------------------------------------------------------------:|:-------------------------------:|:--------------------------------------------------:|:-------------------:|:---------:|:-------------------------------------------------------------------------------:|
| [Polyglot-Ko](https://github.com/EleutherAI/polyglot) | 1.3B, 3.8B, 5.8B, 12.8B | [EleutherAI](https://github.com/EleutherAI) | GPT-NeoX | ✅ | [🤗 Huggingface](https://huggingface.co/EleutherAI/polyglot-ko-12.8b/tree/main) |
| [KoAlpaca](https://github.com/Beomi/KoAlpaca) | 7B, 13B, 30B, 65B / 5.8B, 12.8B | [beomi](https://github.com/Beomi) | llama / Polyglot-Ko | ❌ / ✅ | [🤗 Huggingface](https://huggingface.co/beomi/KoAlpaca-Polyglot-12.8B) |
| [KuLLM](https://github.com/nlpai-lab/KULLM) | 5.8B, 12.8B | [고려대학교](https://github.com/nlpai-lab) | Polyglot-Ko | ✅ | [🤗 Huggingface](https://huggingface.co/nlpai-lab/kullm-polyglot-12.8b-v2) |
| [KORani](https://github.com/krafton-ai/KORani) | 13B / 12.8B | [KRAFTON](https://github.com/krafton-ai) | llama / Polyglot-Ko | ❌ / ✅ | [🤗 Huggingface](https://huggingface.co/KRAFTON/KORani-v3-13B) |
| [K(G)OAT](https://github.com/Marker-Inc-Korea/K-G-OAT) | 5.8B | [Marker Inc.](https://github.com/Marker-Inc-Korea) | Polyglot-ko | ✅ | [🤗 Huggingface](https://huggingface.co/DopeorNope/KOAT-5.8b/tree/main) |
| [KoVicuna](https://github.com/melodysdreamj/KoVicuna) | 7B | [melodysdreamj](https://github.com/melodysdreamj) | Vicuna 7B | ❓ | [🤗 Huggingface](https://huggingface.co/junelee/ko_vicuna_7b) |
| [Kollama](https://huggingface.co/beomi/kollama-33b) | 7B, 13B, 33B | [beomi](https://github.com/Beomi) | llama | ❌ | [🤗 Huggingface](https://huggingface.co/beomi/kollama-33b) |
| [Llama-2-Ko](https://huggingface.co/beomi/llama-2-ko-7b) | 7B, 70B | [beomi](https://github.com/Beomi) | llama-2 | ✅ | [🤗 Huggingface](https://huggingface.co/beomi/llama-2-ko-7b) || [KoLlama2](https://github.com/psymon-dev/KoLlama2) | ㅇ | [psymon-dev](https://github.com/psymon-dev) | llama-2 | ❓ | [🤗 Huggingface](https://huggingface.co/psymon/KoLlama2-7b) |
| [komt](https://github.com/davidkim205/komt) | 7B, 13B | [davidkim205](https://github.com/davidkim205) | llama-2 | ✅ | [🤗 Huggingface](https://huggingface.co/davidkim205/komt-Llama-2-13b-hf) |
| [llama-2-korean](https://huggingface.co/quantumaikr/llama-2-70b-fb16-korean) | 7B, 13B, 70B | [(주)퀀텀아이](https://github.com/quantumaikr) | llama-2 | ✅ | [🤗 Huggingface](https://huggingface.co/quantumaikr/llama-2-70b-fb16-korean) |
| [KoreanLM](https://github.com/quantumaikr/KoreanLM) | 1.5B, 3B | [(주)퀀텀아이](https://github.com/quantumaikr) | ❓ | ✅ | [🤗 Huggingface](https://huggingface.co/quantumaikr/KoreanLM) |
| [KoRMKV](https://huggingface.co/beomi/KoRWKV-1.5B) | 1.5B, 6B | [beomi](https://github.com/Beomi) | RMKVv4 | ❓ | [🤗 Huggingface](https://huggingface.co/beomi/KoRWKV-1.5B) |
| [KoAlpaca-KoRMKV](https://huggingface.co/beomi/KoAlpaca-KoRWKV-6B) | 1.5B, 6B | [beomi](https://github.com/Beomi) | KoRMKV | ❓ | [🤗 Huggingface](https://huggingface.co/beomi/KoRWKV-1.5B) |
| [KoGPT](https://github.com/kakaobrain/kogpt) | 6B | [카카오브레인](https://github.com/kakaobrain) | ❓ | ❌ | [🤗 Huggingface](https://huggingface.co/kakaobrain/kogpt/tree/KoGPT6B-ryan1.5b) |
| [KoGPT2](https://github.com/SKT-AI/KoGPT2) | 1.5B | [SKT](https://github.com/SKT-AI) | GPT-2 | ❌ | ❓ |
| [Llama-2-ko-7b-Chat](https://huggingface.co/kfkas/Llama-2-ko-7b-Chat) | 7B | [taemin6697](https://github.com/taemin6697) | llama-2 | ✅ | [🤗 Huggingface](https://huggingface.co/kfkas/Llama-2-ko-7b-Chat) |
| [42dot_LLM-PLM-1.3B](https://huggingface.co/42dot/42dot_LLM-PLM-1.3B) | 1.3B | [42dot](https://github.com/42dot) | llama-2 | ❌ | [🤗 Huggingface](https://huggingface.co/42dot/42dot_LLM-PLM-1.3B) |
| [42dot_LLM-SFT-1.3B](https://huggingface.co/42dot/42dot_LLM-SFT-1.3B) | 1.3B | [42dot](https://github.com/42dot) | llama-2 | ❌ | [🤗 Huggingface](https://huggingface.co/42dot/42dot_LLM-SFT-1.3B) |
| [Ko-Platypus](https://github.com/Marker-Inc-Korea/KO-Platypus) | 7B | [Marker Inc.](https://github.com/Marker-Inc-Korea) | llama-2 | ❓ | [🤗 Huggingface](https://huggingface.co/kyujinpy/KO-Platypus2-7B-ex) |
| [sitebunny](https://huggingface.co/42MARU/sitebunny-13b) | 13B | [42MARU](https://huggingface.co/42MARU) | llama-2 | ✅ | [🤗 Huggingface](https://huggingface.co/42MARU/sitebunny-13b) |
| [ChatSKKU](https://huggingface.co/jojo0217/ChatSKKU5.8B) | 5.8B | [jojo0217](https://github.com/JoJo0217) | Polyglot-Ko | ❓ | [🤗 Huggingface](https://huggingface.co/jojo0217/ChatSKKU5.8B) |
| [nallm](https://github.com/Nara-Information/NA-LLM) | 1.3B, 3.8B | [나라지식정보](https://github.com/Nara-Information) | Polyglot-Ko | ❓ | [🤗 Huggingface](https://huggingface.co/Nara-Lab/nallm-polyglot-ko-3.8b-base) |
| [Mi:dm (믿:음)](https://huggingface.co/KT-AI/midm-bitext-S-7B-inst-v1) | 7B | [KT](https://huggingface.co/KT-AI) | ❓ | ❌ | [🤗 Huggingface](https://huggingface.co/KT-AI/midm-bitext-S-7B-inst-v1) |
## 한국어 오픈소스 LLM 모델
### 1. Polyglot-Ko 기반
- [Polyglot-Ko](https://github.com/EleutherAI/polyglot)
- [KoAlpaca](https://github.com/Beomi/KoAlpaca)
- [KuLLM](https://github.com/nlpai-lab/KULLM)
- [KORani](https://github.com/krafton-ai/KORani)
- [K(G)OAT](https://github.com/Marker-Inc-Korea/K-G-OAT)
- [ChatSKKU](https://huggingface.co/jojo0217/ChatSKKU5.8B)
- [nallm](https://github.com/Nara-Information/NA-LLM)
### 2-1. Llama 기반
- [Kollama](https://huggingface.co/beomi/kollama-33b)
- [KoAlpaca](https://github.com/Beomi/KoAlpaca)
- [KORani](https://github.com/krafton-ai/KORani)
### 2-2. Llama-2 기반
- [Llama-2-Ko](https://huggingface.co/beomi/llama-2-ko-7b)
- [komt](https://github.com/davidkim205/komt)
- [llama-2-korean](https://huggingface.co/quantumaikr/llama-2-70b-fb16-korean)
- [Llama-2-ko-7b-Chat](https://huggingface.co/kfkas/Llama-2-ko-7b-Chat)
- [42dot_LLM-PLM-1.3B](https://huggingface.co/42dot/42dot_LLM-PLM-1.3B)
- [42dot_LLM-SFT-1.3B](https://huggingface.co/42dot/42dot_LLM-SFT-1.3B)
- [Ko-Platypus](https://github.com/Marker-Inc-Korea/KO-Platypus)
- [sitebunny](https://huggingface.co/42MARU/sitebunny-13b)
### 3. 기타
- [KoVicuna](https://github.com/melodysdreamj/KoVicuna)
- [KoRMKV](https://huggingface.co/beomi/KoRWKV-1.5B)
- [KoAlpaca-KoRMKV](https://huggingface.co/beomi/KoAlpaca-KoRWKV-6B)
- [KoGPT](https://github.com/kakaobrain/kogpt)
- [KoGPT2](https://github.com/SKT-AI/KoGPT2)
- [KoreanLM](https://github.com/quantumaikr/KoreanLM)
- [Mi:dm (믿:음)](https://huggingface.co/KT-AI/midm-bitext-S-7B-inst-v1)
## Contributing
틀린 점이 있거나 새로운 한국어 LLM이 있다면 언제나 PR을 보내주세요!