---
language:
- ko
- en
library_name: transformers
license: cc-by-nc-sa-4.0
pipeline_tag: text-generation
tags:
- pytorch
---

# Model Card for RedWhale-tv-10.8B-v1.0

<!-- Provide a quick summary of what the model is/does. -->

<!--
<img src="https://huggingface.co/TwinDoc/RedWhale-tv-10.8B-v1.0/resolve/main/company_agilesoda__icon.png"  width="648">
-->
<img src="https://huggingface.co/TwinDoc/RedWhale-tv-10.8B-v1.0/resolve/main/company_agilesoda__icon_RWTV.png"  width="648">


## Model Description

**RedWhale**은 전처리한 한국어 Corpus, 특화된 한국어 Tokenizer, 효과적인 Model initialization,  Continuous Multi-Stage Pretraining strategy 등을 갖추고 있습니다. 이러한 접근 방식은 높은 정확도와 이해도를 유지하면서 Computational costs를 줄여 제한된 리소스에서 Pretraining을 가능하게 해줍니다. **RedWhale** 사용을 원하시면 repo access 요청해주세요.

<!-- Provide a longer summary of what this model is. -->

## About the Model

- **Name:** TwinDoc/RedWhale-tv-10.8B-v1.0
- **Foundation Model:** upstage/SOLAR-10.7B-v1.0
- **Train Corpus:** [preprocessed AI-Hub datasets](https://huggingface.co/datasets/TwinDoc/agilesoda-corpus-AIHUB_splited_shffled)
- **Developed by:** 애자일소다 (AGILESODA)
- **Model type:** llama
- **Language(s) (NLP):** 한국어, 영어
- **License:**  cc-by-nc-sa-4.0
- **Paper:** [RedWhale: An Adapted Korean LLM Through Efficient Continual Pretraining
](https://arxiv.org/abs/2408.11294)


<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

## Load the Model

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

```
from transformers import AutoTokenizer
from transformers import AutoModelForCausalLM

YOUR_HF_TOKEN_READ = "hf_..."
model_name_or_path = "TwinDoc/RedWhale-tv-10.8B-v1.0"
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, token=YOUR_HF_TOKEN_READ)
model = AutoModelForCausalLM.from_pretrained(model_name_or_path, token=YOUR_HF_TOKEN_READ)
```

## Generate Text

```
text = "대한민국의 수도는"
encodings = tokenizer(text, return_tensors='pt')
terminators = [tokenizer.eos_token_id] + tokenizer("\n", add_special_tokens=False)["input_ids"]

outputs = model.generate(**encodings, eos_token_id=terminators)
generated_text = tokenizer.batch_decode(outputs)[0]
# '<s> 대한민국의 수도는 서울이다.\n'
```

## License

<img src="https://huggingface.co/TwinDoc/RedWhale-tv-10.8B-v1.0/resolve/main/license__icon.png"  width="324">

The content of this project, created by AGILESODA, is licensed under the [Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)](https://creativecommons.org/licenses/by-nc-sa/4.0/).

## Citation

```
@misc{vo2024redwhaleadaptedkoreanllm,
      title={RedWhale: An Adapted Korean LLM Through Efficient Continual Pretraining}, 
      author={Anh-Dung Vo and Minseong Jung and Wonbeen Lee and Daewoo Choi},
      year={2024},
      eprint={2408.11294},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2408.11294}, 
}
```


**Built with:**

<a href="http://www.agilesoda.com/sub/twin_doc.php">
    <img src="https://huggingface.co/TwinDoc/RedWhale-tv-10.8B-v1.0/resolve/main/company_agilesoda_twindoc__icon.png" alt="AgileSoda TwinDoc Icon">
</a>