JeongwonChoi's picture
Update README.md
6ddc4cf verified
|
raw
history blame
3.24 kB
metadata
tags:
  - text-generation
license: cc-by-nc-4.0
language:
  - ko
base_model: beomi/OPEN-SOLAR-KO-10.7B
pipeline_tag: text-generation

DataVortexS-10.7B-dpo-v1.10

DataVortex

Model Details

Base Model

beomi/OPEN-SOLAR-KO-10.7B

Trained On

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

Instruction format

It follows Alpaca (Chat) format.

E.g.

text = """\
### System:
당신은 μ‚¬λžŒλ“€μ΄ 정보λ₯Ό 찾을 수 μžˆλ„λ‘ λ„μ™€μ£ΌλŠ” 인곡지λŠ₯ λΉ„μ„œμž…λ‹ˆλ‹€.

### User:
λŒ€ν•œλ―Όκ΅­μ˜ μˆ˜λ„λŠ” μ–΄λ””μ•Ό?

### Assistant:
λŒ€ν•œλ―Όκ΅­μ˜ μˆ˜λ„λŠ” μ„œμšΈμž…λ‹ˆλ‹€.

### User:
μ„œμšΈ μΈκ΅¬λŠ” 총 λͺ‡ λͺ…이야?
"""

Model Benchmark

Ko LM Eval Harness

Task 0-shot 5-shot 10-shot 50-shot
kobest_boolq 0.874261 0.897165 0.904985 0.907857
kobest_copa 0.807479 0.845701 0.860809 0.8719
kobest_hellaswag 0.504865 0.502074 0.50717 0.51609
kobest_sentineg 0.409404 0.967251 0.992443 0.982367
Average 0.649002 0.803048 0.816352 0.819553

Ko-LLM-Leaderboard

Average Ko-ARC Ko-HellaSwag Ko-MMLU Ko-TruthfulQA Ko-CommonGen V2
56.32 54.27 63.16 49.95 55.08 59.15

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-v1.10")
tokenizer = AutoTokenizer.from_pretrained("Edentns/DataVortexS-10.7B-dpo-v1.10")

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

This model is licensed under the cc-by-nc-4.0. which allows others to share and adapt the model for non-commercial purposes.