README.md

---
language:
- ko
datasets:
- kyujinpy/KoCoT_2000
library_name: transformers
pipeline_tag: text-generation
license: cc-by-nc-sa-4.0
---
**(주)미디어그룹사람과숲과 (주)마커의 LLM 연구 컨소시엄에서 개발된 모델입니다**  
**The license is `cc-by-nc-sa-4.0`.**  

# **KoT-platypus2**  
![img](./KoT-platypus2.png)  
**CoT + KO-platypus2 = KoT-platypus2**

## Model Details

**Model Developers** Kyujin Han (kyujinpy)

**Input** Models input text only.

**Output** Models generate text only.

**Model Architecture**  
KoT-platypus2-13B is an auto-regressive language model based on the LLaMA2 transformer architecture.

**Repo Link**  
Github KoT-platypus: [KoT-platypus2](https://github.com/KyujinHan/KoT-platypus)  

**Base Model**  
[KO-Platypus2-13B](https://huggingface.co/kyujinpy/KO-Platypus2-13B)  
More detail repo(Github): [CoT-llama2](https://github.com/Marker-Inc-Korea/CoT-llama2)  
More detail repo(Github): [KO-Platypus2](https://github.com/Marker-Inc-Korea/KO-Platypus)    

**Training Dataset**  
I use [KoCoT_2000](https://huggingface.co/datasets/kyujinpy/KoCoT_2000).  
Using DeepL, translate about [kaist-CoT](https://huggingface.co/datasets/kaist-ai/CoT-Collection).

I use A100 GPU 40GB and COLAB, when trianing.

**Training Hyperparameters**
| Hyperparameters | Value |  
| --- | --- |  
| batch_size | `64` |   
| micro_batch_size | `1` |  
| Epochs | `15` |  
| learning_rate | `1e-5` |  
| cutoff_len | `4096` |  
| lr_scheduler | `linear` |  
| base_model | `kyujinpy/KO-Platypus2-13B` |  


# **Model Benchmark**

## KO-LLM leaderboard
- Follow up as [Open KO-LLM LeaderBoard](https://huggingface.co/spaces/upstage/open-ko-llm-leaderboard).  
  
![img](./leaderboard.png)
| Model | Average |Ko-ARC | Ko-HellaSwag | Ko-MMLU | Ko-TruthfulQA | Ko-CommonGen V2 |
| --- | --- | --- | --- | --- | --- | --- |
|KoT-Platypus2-13B(ours) | 49.55 | 43.69 | 53.05 | 42.29 | 43.34 | 65.38 |  
| [KO-Platypus2-13B](https://huggingface.co/kyujinpy/KO-Platypus2-13B) | 47.90 | 44.20 | 54.31 | 42.47 | 44.41 | 54.11 | 
| [hyunseoki/ko-en-llama2-13b](https://huggingface.co/hyunseoki/ko-en-llama2-13b) | 46.68 | 42.15 | 54.23 | 38.90 | 40.74 | 57.39 |
| [MarkrAI/kyujin-CoTy-platypus-ko-12.8b](https://huggingface.co/MarkrAI/kyujin-CoTy-platypus-ko-12.8b) | 46.44 | 34.98 | 49.11 | 25.68 | 37.59 | 84.86 | 
| [momo/polyglot-ko-12.8b-Chat-QLoRA-Merge](https://huggingface.co/momo/polyglot-ko-12.8b-Chat-QLoRA-Merge) | 45.71 | 35.49 | 49.93 | 25.97 | 39.43 | 77.70 |   
> Compare with Top 4 SOTA models. (update: 10/07)

  
# Implementation Code
```python
### KO-Platypus
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

repo = "kyujinpy/KoT-platypus2-13B"
CoT-llama = AutoModelForCausalLM.from_pretrained(
        repo,
        return_dict=True,
        torch_dtype=torch.float16,
        device_map='auto'
)
CoT-llama_tokenizer = AutoTokenizer.from_pretrained(repo)
```

> Readme format: [beomi/llama-2-ko-7b](https://huggingface.co/beomi/llama-2-ko-7b)

---