Instructions to use Minwoo-Study/AIHUB_IndonesianCorpus_Generation with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Minwoo-Study/AIHUB_IndonesianCorpus_Generation with Transformers:

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

pipe = pipeline("text-generation", model="Minwoo-Study/AIHUB_IndonesianCorpus_Generation")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Minwoo-Study/AIHUB_IndonesianCorpus_Generation")
model = AutoModelForCausalLM.from_pretrained("Minwoo-Study/AIHUB_IndonesianCorpus_Generation", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Minwoo-Study/AIHUB_IndonesianCorpus_Generation with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Minwoo-Study/AIHUB_IndonesianCorpus_Generation"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Minwoo-Study/AIHUB_IndonesianCorpus_Generation",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Minwoo-Study/AIHUB_IndonesianCorpus_Generation

SGLang

How to use Minwoo-Study/AIHUB_IndonesianCorpus_Generation 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 "Minwoo-Study/AIHUB_IndonesianCorpus_Generation" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Minwoo-Study/AIHUB_IndonesianCorpus_Generation",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "Minwoo-Study/AIHUB_IndonesianCorpus_Generation" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Minwoo-Study/AIHUB_IndonesianCorpus_Generation",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Minwoo-Study/AIHUB_IndonesianCorpus_Generation with Docker Model Runner:
```
docker model run hf.co/Minwoo-Study/AIHUB_IndonesianCorpus_Generation
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

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

Check out the documentation for more information.

99. 인도네시아어 말뭉치 데이터 -2. 언어생성모델

모델 설명

2023년 인공지능 학습용 데이터 구축사업으로 구축한 2억 토큰 규모의 인도네시아어 말뭉치 데이터로 초거대 AI 언어생성모델 개발

모델 아키텍처

META AI에서 개발한 다국어 모델 LLaMA(Large Language Model Meta AI) 중 LLaMA-2-7b-chat(https://huggingface.co/meta-llama/Llama-2-7b-chat-hf) 활용

Unsupervised Finetuning

말뭉치 전체로 LoRA(Low Rank Adaptation)을 활용한 Unsupervised Finetuing 진행

Supervised Finetuning

사전에 제작한 한국 관련 질문 답변셋을 통한 Supervised Finetuing 진행

데이터 설명

Input

인도네시아어로 모델에 질문 혹은 명령 가능 ex :
Siapa sajakah penulis terkenal dari Korea? 한국의 유명한 작가는 누구인가요?

Output

인도네시아어 답변 생성 ex : Penulis terkenal Korea: Kim Eun-sook, Park Ji-eun. 한국의 유명 작가: 김은숙, 박지은.

학습 파라미터

Unsupervised Finetuing / Supervised Finetuning 동일

    --optim 'adamw_torch_fused' \
    --seed 0 --data_seed 0 \
    --dataloader_num_workers 30 \
    --per_device_train_batch_size 2 --per_device_eval_batch_size 2 \
    --low_cpu_mem_usage false \
    --evaluation_strategy "steps" --eval_steps 5000 \
    --save_strategy "steps" --save_steps 5000 \
    --save_total_limit 2 \
    --gradient_accumulation_steps 4 \
    --learning_rate 3e-4 \
    --lr_scheduler_type "cosine" \
    --warmup_steps 28 \
    --num_train_epochs 5 \
    --use_lora \
    --lora_rank 64 \
    --lora_alpha 16 \
    --lora_dropout 0.05 \

평가 기준

Human Evaluation

생성형 모델 6개 중 비교모델 5개 선정 (ChatGPT 3.5-Turbo, IndoGPT, BLOOMZ, mT0)
학습된 모델을 포함하여 5개 후보군 모델이 생성한 각 답변에 대해 3명의 평가자가 평가한 평균 점수로 종합 순위를 판단함.
평가자에게는 모델 정보가 제공되지 않음.

목표 순위 3위 / 최종 순위 2위

저작권 라이센스 표기

MIT License

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Downloads last month: 9

Model tree for Minwoo-Study/AIHUB_IndonesianCorpus_Generation

Quantizations

1 model