Instructions to use UICHEOL-HWANG/KoGPT2-fairytail with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use UICHEOL-HWANG/KoGPT2-fairytail with Transformers:

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

pipe = pipeline("text-generation", model="UICHEOL-HWANG/KoGPT2-fairytail")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("UICHEOL-HWANG/KoGPT2-fairytail")
model = AutoModelForCausalLM.from_pretrained("UICHEOL-HWANG/KoGPT2-fairytail", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use UICHEOL-HWANG/KoGPT2-fairytail with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "UICHEOL-HWANG/KoGPT2-fairytail"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "UICHEOL-HWANG/KoGPT2-fairytail",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/UICHEOL-HWANG/KoGPT2-fairytail

SGLang

How to use UICHEOL-HWANG/KoGPT2-fairytail 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 "UICHEOL-HWANG/KoGPT2-fairytail" \
    --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": "UICHEOL-HWANG/KoGPT2-fairytail",
		"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 "UICHEOL-HWANG/KoGPT2-fairytail" \
        --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": "UICHEOL-HWANG/KoGPT2-fairytail",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use UICHEOL-HWANG/KoGPT2-fairytail with Docker Model Runner:
```
docker model run hf.co/UICHEOL-HWANG/KoGPT2-fairytail
```

KoGPT2 fine Tuned model

타겟 : 동화 데이터를 통한 동화 생성 모델
base model : "skt/kogpt2-base-v2"
batch-size : 8
max_length : 512
num_epochs : 5
learning_rate : 5e-5

특이점

동화 데이터 특성상 한 편당 길이가 매우 길어 chunk 단위로 짜른 뒤 데이터를 학습 시킴
- 그렇게 나눈 덕에 학습 데이터가 매우 많아 VRAM 과부하 발생
  - frac=0.3 비율로 짤라서 학습
  - 추후 인퍼런스 예정
- github
https://github.com/UICHEOL-HWANG/KoGPT2-Generate_fairy.git

Downloads last month: 8

Safetensors

Model size

0.1B params

Tensor type

F32

Model tree for UICHEOL-HWANG/KoGPT2-fairytail

Base model

skt/kogpt2-base-v2

Finetuned

(20)

this model

UICHEOL-HWANG
/

KoGPT2-fairytail

KoGPT2 fine Tuned model

특이점

Model tree for UICHEOL-HWANG/KoGPT2-fairytail

Dataset used to train UICHEOL-HWANG/KoGPT2-fairytail