Instructions to use KRAFTON/Raon-1-mini-preview with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KRAFTON/Raon-1-mini-preview with Transformers:

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

pipe = pipeline("text-generation", model="KRAFTON/Raon-1-mini-preview", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("KRAFTON/Raon-1-mini-preview", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use KRAFTON/Raon-1-mini-preview with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KRAFTON/Raon-1-mini-preview"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KRAFTON/Raon-1-mini-preview",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KRAFTON/Raon-1-mini-preview

SGLang

How to use KRAFTON/Raon-1-mini-preview 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 "KRAFTON/Raon-1-mini-preview" \
    --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": "KRAFTON/Raon-1-mini-preview",
		"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 "KRAFTON/Raon-1-mini-preview" \
        --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": "KRAFTON/Raon-1-mini-preview",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use KRAFTON/Raon-1-mini-preview with Docker Model Runner:
```
docker model run hf.co/KRAFTON/Raon-1-mini-preview
```

Raon-mHC VS Deepseek v4.1's Single-pass mHC ?

by Mem-Mori - opened 8 days ago

Discussion

Mem-Mori

8 days ago

Just curious. Can you tell me what is the main difference between deepseek's single-pass mHC?
I remember deepseek also did a optimation? on the mHC(Tell me if im incorrect)
Raon-mHC는 딥시크 v4.1의 single pass mHC와 어떻게 다른지 여쭤봐도 될까여..? 딥시크도 이번 v4.1 flash 모델에 하드웨어 최적화? 비슷한걸 한걸로 기억해서요!
KDA와 비슷하지만 지우기,쓰기 분리 KDA?로 시작해서 학습하며 분리하는 RDA도 흥미롭네요! Tech report 기대할게요.
한국에도 큰 아키텍처 변화를 시도 + 게이밍 글카/컴에서도 돌아가는 모델을 만드는 회사가..!!
D-22

jihuny

KRAFTON org 8 days ago

Thanks for the interest! You remember correctly.
DeepSeek-V4.1's Single-Pass mHC is an efficiency change. Each block reuses the input-mixing coefficients from the previous block, so the residual stream is read and written only once, which cuts memory traffic.

Raon-mHC changes how the stream-mixing matrix itself is built. DeepSeek approximates it with Sinkhorn iterations while Raon-mHC builds an exactly doubly stochastic matrix in one shot with no iterations (via parametrization). That makes it faster and lighter on memory.

In short, DeepSeek changes when the coefficients are used, and Raon-mHC changes how the mixing matrix is built.
The two are independent and could be combined.

Glad you found RDA interesting too. More details will be in the tech report!

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment