README.md

metadata

tags:
  - text-generation
license: cc-by-nc-sa-4.0
language:
  - ko
base_model: LDCC/LDCC-SOLAR-10.7B
pipeline_tag: text-generation
datasets:
  - mncai/orca_dpo_pairs_ko
  - Ja-ck/Orca-DPO-Pairs-KO
  - We-Want-GPU/Yi-Ko-DPO-Orca-DPO-Pairs

DataVortexS-10.7B-dpo-v0.1

Our Team

Research & Engineering	Product Management
Kwangseok Yang	Seunghyun Choi
Jeongwon Choi	Hyoseok Choi

Model Details

Base Model

LDCC/LDCC-SOLAR-10.7B

Trained On

• OS: Ubuntu 20.04
• GPU: H100 80GB 2ea
• transformers: v4.36.2

Dataset

• mncai/orca_dpo_pairs_ko
• Ja-ck/Orca-DPO-Pairs-KO
• We-Want-GPU/Yi-Ko-DPO-Orca-DPO-Pairs

Instruction format

It follows Alpaca format.

E.g.

text = """\
당신은 사람들이 정보를 찾을 수 있도록 도와주는 인공지능 비서입니다.

### User:
대한민국의 수도는 어디야?

### Assistant:
대한민국의 수도는 서울입니다.

### User:
서울 인구는 총 몇 명이야?
"""

Model Benchmark

Ko LM Eval Harness

Task	0-shot	5-shot	10-shot	50-shot
kobest_boolq	0.334282	0.891367	0.896755	0.884441
kobest_copa	0.697763	0.716762	0.724769	0.751746
kobest_hellaswag	0.432047	0.458301	0.443993	0.458232
kobest_sentineg	0.49353	0.954657	0.964735	0.949606
Average	0.4894055	0.75527175	0.757563	0.76100625

Ko-LLM-Leaderboard

Average	Ko-ARC	Ko-HellaSwag	Ko-MMLU	Ko-TruthfulQA	Ko-CommonGen V2
53.21	47.87	57.18	54.82	53.64	52.54

Implementation Code

This model contains the chat_template instruction format.
You can use the code below.

from transformers import AutoModelForCausalLM, AutoTokenizer

device = "cuda" # the device to load the model onto

model = AutoModelForCausalLM.from_pretrained("Edentns/DataVortexS-10.7B-dpo-v0.1")
tokenizer = AutoTokenizer.from_pretrained("Edentns/DataVortexS-10.7B-dpo-v0.1")

messages = [
    {"role": "system", "content": "당신은 사람들이 정보를 찾을 수 있도록 도와주는 인공지능 비서입니다."},
    {"role": "user", "content": "대한민국의 수도는 어디야?"},
    {"role": "assistant", "content": "대한민국의 수도는 서울입니다."},
    {"role": "user", "content": "서울 인구는 총 몇 명이야?"}
]

encodeds = tokenizer.apply_chat_template(messages, return_tensors="pt")

model_inputs = encodeds.to(device)
model.to(device)

generated_ids = model.generate(model_inputs, max_new_tokens=1000, do_sample=True)
decoded = tokenizer.batch_decode(generated_ids)
print(decoded[0])

License

The model is licensed under the cc-by-nc-sa-4.0 license, which allows others to copy, modify, and share the work non-commercially, as long as they give appropriate credit and distribute any derivative works under the same license.