Instructions to use MDDDDR/Ko-Luxia-8B-it-v0.2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MDDDDR/Ko-Luxia-8B-it-v0.2 with Transformers:

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

pipe = pipeline("text-generation", model="MDDDDR/Ko-Luxia-8B-it-v0.2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MDDDDR/Ko-Luxia-8B-it-v0.2")
model = AutoModelForCausalLM.from_pretrained("MDDDDR/Ko-Luxia-8B-it-v0.2", 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 MDDDDR/Ko-Luxia-8B-it-v0.2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MDDDDR/Ko-Luxia-8B-it-v0.2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MDDDDR/Ko-Luxia-8B-it-v0.2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/MDDDDR/Ko-Luxia-8B-it-v0.2

SGLang

How to use MDDDDR/Ko-Luxia-8B-it-v0.2 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 "MDDDDR/Ko-Luxia-8B-it-v0.2" \
    --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": "MDDDDR/Ko-Luxia-8B-it-v0.2",
		"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 "MDDDDR/Ko-Luxia-8B-it-v0.2" \
        --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": "MDDDDR/Ko-Luxia-8B-it-v0.2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use MDDDDR/Ko-Luxia-8B-it-v0.2 with Docker Model Runner:
```
docker model run hf.co/MDDDDR/Ko-Luxia-8B-it-v0.2
```

Model Card for Model ID

base_model : Ko-Llama3-Luxia-8B

Basic usage

# pip install accelerate
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

tokenizer = AutoTokenizer.from_pretrained("MDDDDR/Ko-Luxia-8B-it-v0.2")
model = AutoModelForCausalLM.from_pretrained(
    "MDDDDR/Ko-Luxia-8B-it-v0.2",
    device_map="auto",
    torch_dtype=torch.bfloat16
)

input_text = "사과가 뭐야?"
input_ids = tokenizer(input_text, return_tensors="pt").to("cuda")

outputs = model.generate(**input_ids)
print(tokenizer.decode(outputs[0]))

Training dataset

dataset : kyujinpy/KOpen-platypus

lora_config and bnb_config in Training

bnd_config = BitsAndBytesConfig(
  load_in_4bit = True
)

lora_config = LoraConfig(
  r = 16,
  lora_alpha = 16,
  lora_dropout = 0.05,
  target_modules = ['gate_proj', 'up_proj', 'down_proj']
)

Hardware

RTX 3090 Ti 24GB x 1

Model Benchmark Results

Tasks	Version	Filter	Metric		Value		Stderr
kobest_boolq	1	none	acc	↑	0.5278	±	0.0133
		none	f1	↑	0.3954	±	N/A
kobest_copa	1	none	acc	↑	0.7380	±	0.0139
		none	f1	↑	0.7372	±	N/A
kobest_hellaswag	1	none	acc	↑	0.4800	±	0.0224
		none	acc_norm	↑	0.6180	±	0.0218
		none	f1	↑	0.4774	±	N/A
kobest_sentineg	1	none	acc	↑	0.5390	±	0.0250
		none	f1	↑	0.5037	±	N/A

Downloads last month: 4

Safetensors

Model size

8B params

Tensor type

F16

MDDDDR
/

Ko-Luxia-8B-it-v0.2