JeongwonChoi's picture
Update README.md
6d8dbec verified
|
raw
history blame
3.65 kB
metadata
tags:
  - text-generation
license: cc-by-nc-sa-4.0
language:
  - ko
base_model: megastudy/M-SOLAR-10.7B-v1.3
pipeline_tag: text-generation

DataVortexS-10.7B-dpo-v1.5

DataVortex

Our Team

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

Model Details

Base Model

megastudy/M-SOLAR-10.7B-v1.3

Trained On

  • OS: Ubuntu 22.04
  • GPU: H100 80GB 4ea
  • transformers: v4.36.2

Instruction format

It follows ChatML format.

E.g.

text = """\
<|im_start|>system
당신은 μ‚¬λžŒλ“€μ΄ 정보λ₯Ό 찾을 수 μžˆλ„λ‘ λ„μ™€μ£ΌλŠ” 인곡지λŠ₯ λΉ„μ„œμž…λ‹ˆλ‹€.<|im_end|>
<|im_start|>user
λŒ€ν•œλ―Όκ΅­μ˜ μˆ˜λ„λŠ” μ–΄λ””μ•Ό?<|im_end|>
<|im_start|>assistant
λŒ€ν•œλ―Όκ΅­μ˜ μˆ˜λ„λŠ” μ„œμšΈμž…λ‹ˆλ‹€.<|im_end|>
<|im_start|>user
μ„œμšΈ μΈκ΅¬λŠ” 총 λͺ‡ λͺ…이야?<|im_end|>
<|im_start|>assistant
"""

Model Benchmark

Ko LM Eval Harness

Task 0-shot 5-shot 10-shot 50-shot
kobest_boolq 0.34687 0.930158 0.943013 0.938029
kobest_copa 0.693351 0.751805 0.75772 0.771704
kobest_hellaswag 0.480736 0.470852 0.474766 0.478576
kobest_sentineg 0.789423 0.962208 0.967241 0.964717
Average 0.577595 0.778756 0.785685 0.788257

Ko-LLM-Leaderboard

Average Ko-ARC Ko-HellaSwag Ko-MMLU Ko-TruthfulQA Ko-CommonGen V2
55.32 52.13 61.27 53.99 49.71 59.5

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-v1.5")
tokenizer = AutoTokenizer.from_pretrained("Edentns/DataVortexS-10.7B-dpo-v1.5")

messages = [
    {"role": "system", "content": "당신은 μ‚¬λžŒλ“€μ΄ 정보λ₯Ό 찾을 수 μžˆλ„λ‘ λ„μ™€μ£ΌλŠ” 인곡지λŠ₯ λΉ„μ„œμž…λ‹ˆλ‹€."},
    {"role": "user", "content": "λŒ€ν•œλ―Όκ΅­μ˜ μˆ˜λ„λŠ” μ–΄λ””μ•Ό?"},
    {"role": "assistant", "content": "λŒ€ν•œλ―Όκ΅­μ˜ μˆ˜λ„λŠ” μ„œμšΈμž…λ‹ˆλ‹€."},
    {"role": "user", "content": "μ„œμšΈ μΈκ΅¬λŠ” 총 λͺ‡ λͺ…이야?"}
]

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.