File size: 1,557 Bytes
409affd
 
 
 
 
 
 
 
 
90763f4
d3f144a
90763f4
d3f144a
 
 
 
90763f4
 
 
 
 
8079260
c0dff62
90763f4
 
 
 
 
 
 
852a7f1
a93d7e4
efd9567
921e5c6
90763f4
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
---
title: README
emoji: 👀
colorFrom: green
colorTo: blue
sdk: static
pinned: false
---

<h1 align="center">KORMo: Korean Open Reasoning Model for Everyone</h1>
<p align="center">
  An open-source hub for Korean language data and model research  
</p>

---


## 🧠 Open Models

- **KORMo-Team/KORMo-tokenizer** — A tokenizer optimized for bilingual (Korean–English) language representation  
- **KORMo-Team/KORMo-10B-base** — The <b>KORMo-10B</b> pretrained model trained on large-scale Korean and English corpora  
- **KORMo-Team/KORMo-10B-sft** — A fine-tuned model enhanced with long-context reasoning and instruction-following data
- **KORMo-Team/KORMo-10B-inst** — Final instruction-tuned model with reasoning enhancement and RL (Coming soon; currently awaiting GPU availability)

> 💡 You can explore the full training history and checkpoints in each model’s **`Revisions` tab** on Hugging Face.


---

## 🌐 Links
- **Technical Report** — https://arxiv.org/pdf/2510.09426
- **Technical Report(Slide-Korean)** — https://github.com/MLP-Lab/KORMo-tutorial/blob/main/20251009_MLP_KORMo(Korean).pdf
- **Tutorial on Github** — https://github.com/MLP-Lab/KORMo-tutorial
- **Tutorial on youtube** — https://www.youtube.com/@MLPLab

---

### 📖 About KORMo

KORMo is an open research initiative dedicated to advancing Korean language understanding and generation through large-scale, fully open-source models and datasets.  
We aim to make Korean NLP research transparent, reproducible, and accessible to the global community.