cpm-ai/Llama3-Ocelot-8B-instruct-v01
030
<p align="left"> <img src="https://huggingface.co/cpm-ai/Ocelot-Ko-self-instruction-10.8B-v1.0/resolve/main/ocelot.webp" width="50%"/> <p>
Kor-LLAMA3 Model
Update @ 2024.06.05: First release of Llama3-Ocelot-8B-instruct-v01
<!-- Provide a quick summary of what the model is/does. -->
This model card corresponds to the 8B Instruct version of the Llama-Ko model.
The train wad done on A100-80GB
Resources and Technical Documentation:
- llama Model
- Orca-Math
- ko_Ultrafeedback_binarized
Citation
Model Developers: frcp, nebchi, pepperonipizza97
Model Information
It is an LLM model capable of generating Korean text, trained on a pre-trained base model with high-quality Korean SFT dataset and DPO dataset.
Inputs and outputs
- Input: Text string, such as a question, a prompt, or a document to be summarized.
- Output: Generated Korean-language text in response to the input, such as an answer to a question, or a summary of a document.
Running the model on a single / multi GPU
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("cpm-ai/Ocelot-Ko-self-instruction-10.8B-v1.0")
model = AutoModelForCausalLM.from_pretrained("cpm-ai/Ocelot-Ko-self-instruction-10.8B-v1.0", device_map="auto")
pipe = pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=4096, streamer=streamer)
text = '대한민국의 수도는 어디인가요?'
messages = [
{
"role": "user",
"content": "{}".format(text)
}
]
prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipe(
prompt,
temperature=0.2,
add_special_tokens=True
)
print(outputs[0]["generated_text"][len(prompt):])
results
대한민국의 수도는 서울특별시입니다.
서울특별시에는 청와대, 국회의사당, 대법원 등 대한민국의 주요 정부기관이 위치해 있습니다.
또한 서울시는 대한민국의 경제, 문화, 교육, 교통의 중심지로써 대한민국의 수도이자 대표 도시입니다.제가 도움이 되었길 바랍니다. 더 궁금한 점이 있으시면 언제든지 물어보세요!@misc {cpm-ai/Ocelot-Ko-self-instruction-10.8B-v1.0,
author = { {frcp, nebchi, pepperonipizza97} },
title = { solar-kor-resume},
year = 2024,
url = { https://huggingface.co/cpm-ai/Ocelot-Ko-self-instruction-10.8B-v1.0 },
publisher = { Hugging Face }
}Results in LogicKor* are as follows: