Instructions to use splo2t/deft-korean-alpaca with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use splo2t/deft-korean-alpaca with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="splo2t/deft-korean-alpaca")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("splo2t/deft-korean-alpaca")
model = AutoModelForCausalLM.from_pretrained("splo2t/deft-korean-alpaca", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use splo2t/deft-korean-alpaca with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "splo2t/deft-korean-alpaca"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "splo2t/deft-korean-alpaca",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/splo2t/deft-korean-alpaca

SGLang

How to use splo2t/deft-korean-alpaca with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "splo2t/deft-korean-alpaca" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "splo2t/deft-korean-alpaca",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "splo2t/deft-korean-alpaca" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "splo2t/deft-korean-alpaca",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use splo2t/deft-korean-alpaca with Docker Model Runner:
```
docker model run hf.co/splo2t/deft-korean-alpaca
```

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

DEFT: Korean Alpaca Model (Kogpt-Version)

카카오브레인에서 공개한 한국어 GPT인 Kogpt 모델을 기반으로 Stanford Alpaca의 학습 방법 + LoRA 기법과 KoAlpaca의 데이터셋으로 학습한 한국어 Alpaca모델 DEFT입니다.

hugginface의 peft를 사용해 학습하고자 했으나, 아직 Embedding Layer단의 LoRA 기법 적용이 불가능하여 부득이하게 따로 구현한 방식을 사용하였습니다.

따라서 모델을 Huggingface에서 다운 받으신 후 본 모델 github(https://github.com/Data-Intelligence-Lab/DEFT-korean-alpaca)의 test-deft-chat.ipynb를 참고하여 사용하시기 바랍니다.

현재 모델은 한국어 전용이며 Multi-Turn이 불가능하고, 공격적, 차별적인 언어와 같은 사회적으로 통념되지 않은 표현에 대한 처리가 되어있지 않습니다.

Downloads last month: 28