๐Ÿค– hyperclovax-sft-1.5b-v1

HyperCLOVAX-SEED-Text-Instruct-1.5B ๋ชจ๋ธ์„ ๊ธฐ๋ฐ˜์œผ๋กœ, LoRA ๋ฐฉ์‹์˜ Instruction tuning์„ ์ ์šฉํ•œ ํ•œ๊ตญ์–ด ์–ธ์–ด๋ชจ๋ธ์ž…๋‹ˆ๋‹ค.

์ด ๋ชจ๋ธ์€ ๋‹ค์Œ๊ณผ ๊ฐ™์€ ์กฐ๊ฑด์—์„œ ํ•™์Šต๋˜์—ˆ์Šต๋‹ˆ๋‹ค:

  • Base model: naver-hyperclovax/HyperCLOVAX-SEED-Text-Instruct-1.5B
  • Tuning method: PEFT (LoRA), SFTTrainer
  • ์‚ฌ์šฉ ์˜ˆ์‹œ: ๋ฌธ์„œ ๊ธฐ๋ฐ˜ ์งˆ๋ฌธ ์‘๋‹ต, ํ”„๋กœ์ ํŠธ ์†Œ๊ฐœ ์š”์•ฝ, ๊ฐ„๊ฒฐํ•œ ๋Œ€๋‹ต ์ƒ์„ฑ ๋“ฑ

๐Ÿ”ง ์‚ฌ์šฉ๋ฒ•

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

# ํ† ํฌ๋‚˜์ด์ € ๋กœ๋“œ
tokenizer = AutoTokenizer.from_pretrained("sunnyanna/hyperclovax-sft-1.5b-v1")

# base + LoRA merge
base_model = AutoModelForCausalLM.from_pretrained(
    "naver-hyperclovax/HyperCLOVAX-SEED-Text-Instruct-1.5B"
)
model = PeftModel.from_pretrained(base_model, "sunnyanna/hyperclovax-sft-1.5b-v1")
model = model.merge_and_unload()
model.eval()

# Model Card for hyperclovax_sft_results

This model is a fine-tuned version of [naver-hyperclovax/HyperCLOVAX-SEED-Text-Instruct-1.5B](https://huggingface.co/naver-hyperclovax/HyperCLOVAX-SEED-Text-Instruct-1.5B).
It has been trained using [TRL](https://github.com/huggingface/trl).

## Quick start

```python
from transformers import pipeline

question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
generator = pipeline("text-generation", model="None", device="cuda")
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
print(output["generated_text"])

์–˜์‹œ

๋ฌธ์„œ๋ฅผ ์ฝ๊ณ  ์•„๋ž˜ ์งˆ๋ฌธ์— ๋‹ตํ•˜์„ธ์š”.

### Input:
๋ฌธ์„œ:
Pumati๋Š” ์นด์นด์˜คํ…Œํฌ ๋ถ€ํŠธ์บ ํ”„ ๊ต์œก์ƒ๋“ค์„ ์œ„ํ•œ ํŠธ๋ž˜ํ”ฝ ํ’ˆ์•—์ด ํ”Œ๋žซํผ์ž…๋‹ˆ๋‹ค.  
์‚ฌ์šฉ์ž๊ฐ€ ๋‹ค๋ฅธ ํŒ€ ํ”„๋กœ์ ํŠธ๋ฅผ ์ง์ ‘ ์‚ฌ์šฉํ•˜๊ณ  ๋ฆฌ๋ทฐํ•˜๋ฉด ์šฐ๋ฆฌ ํŒ€์˜ ๋…ธ์ถœ ์ˆœ์œ„๊ฐ€ ์˜ฌ๋ผ๊ฐ€๋Š” ๊ตฌ์กฐ์ž…๋‹ˆ๋‹ค.  

์งˆ๋ฌธ: ์ด๊ฑด ์–ด๋–ค ํ”„๋กœ์ ํŠธ์•ผ?

### Output:
Pumati๋Š” ์„œ๋กœ์˜ ํ”„๋กœ์ ํŠธ๋ฅผ ์ฒดํ—˜ํ•˜๊ณ  ์‘์›ํ•˜๋ฉฐ, ํŠธ๋ž˜ํ”ฝ์„ ๊ณต์œ ํ•  ์ˆ˜ ์žˆ๋„๋ก ์„ค๊ณ„๋œ ๊ต์œก์ƒ ๊ฐ„์˜ ์ปค๋ฎค๋‹ˆํ‹ฐ ๊ธฐ๋ฐ˜ ํ”Œ๋žซํผ์ž…๋‹ˆ๋‹ค.

Training procedure

This model was trained with SFT.

Framework versions

  • PEFT 0.15.2
  • TRL: 0.19.0
  • Transformers: 4.52.4
  • Pytorch: 2.6.0+cu124
  • Datasets: 3.6.0
  • Tokenizers: 0.21.2

Citations

Cite TRL as:

@misc{vonwerra2022trl,
    title        = {{TRL: Transformer Reinforcement Learning}},
    author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
    year         = 2020,
    journal      = {GitHub repository},
    publisher    = {GitHub},
    howpublished = {\url{https://github.com/huggingface/trl}}
}
Downloads last month
11
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for sunnyanna/hyperclovax-sft-1.5b-v1