Instructions to use win10/MagKr-3.2-24B-thinking with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use win10/MagKr-3.2-24B-thinking with Transformers:

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

pipe = pipeline("text-generation", model="win10/MagKr-3.2-24B-thinking")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("win10/MagKr-3.2-24B-thinking")
model = AutoModelForMultimodalLM.from_pretrained("win10/MagKr-3.2-24B-thinking", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use win10/MagKr-3.2-24B-thinking with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "win10/MagKr-3.2-24B-thinking"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "win10/MagKr-3.2-24B-thinking",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/win10/MagKr-3.2-24B-thinking

SGLang

How to use win10/MagKr-3.2-24B-thinking 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 "win10/MagKr-3.2-24B-thinking" \
    --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": "win10/MagKr-3.2-24B-thinking",
		"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 "win10/MagKr-3.2-24B-thinking" \
        --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": "win10/MagKr-3.2-24B-thinking",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use win10/MagKr-3.2-24B-thinking with Docker Model Runner:
```
docker model run hf.co/win10/MagKr-3.2-24B-thinking
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

MMLU-PRO: 總分:63.08

根據我們的評估扣除掉部分因為思維鏈導致的格式錯誤，無法判別外，正確率與原始模型相比更高。

PS:

由於我們不知道原始模型如何評估，所以我們重新檢查了，原始模型得分，發現性能比之原始模型更強，根據我們的測試開啟思考模式下，分數比之原版高上三分達到70以上

原版採用(5-shot CoT)評估，我們僅評估zero-shot

Downloads last month: -

Safetensors

Model size

24B params

Tensor type

BF16

Model tree for win10/MagKr-3.2-24B-thinking

Base model

mistralai/Mistral-Small-3.1-24B-Base-2503

Finetuned

mistralai/Mistral-Small-3.2-24B-Instruct-2506

Finetuned

(65)

this model