Instructions to use win10/K2-nano-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use win10/K2-nano-v1 with Transformers:

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

pipe = pipeline("image-text-to-text", model="win10/K2-nano-v1")
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/K2-nano-v1")
model = AutoModelForMultimodalLM.from_pretrained("win10/K2-nano-v1")
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/K2-nano-v1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "win10/K2-nano-v1"
# 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/K2-nano-v1",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/win10/K2-nano-v1

SGLang

How to use win10/K2-nano-v1 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/K2-nano-v1" \
    --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/K2-nano-v1",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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/K2-nano-v1" \
        --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/K2-nano-v1",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use win10/K2-nano-v1 with Docker Model Runner:
```
docker model run hf.co/win10/K2-nano-v1
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

output002

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the Tri-Ring merge method using /mnt/e/text-generation-webui-1.14/user_data/models/k2-nano-zero as a base.

Models Merged

The following models were included in the merge:

/mnt/f/mergekit-my-test-gemma4/output000
/mnt/f/mergekit-my-test-gemma4/output001
/mnt/e/text-generation-webui-1.14/user_data/models/k2-nano-karcher-100000

Configuration

The following YAML configuration was used to produce this model:

# Tri-Ring: exactly three non-base models are used as the three ring centres.
merge_method: tri_ring
base_model: /mnt/e/text-generation-webui-1.14/user_data/models/k2-nano-zero
models:
  - model: /mnt/f/mergekit-my-test-gemma4/output000
    parameters:
      weight: 0.34
  - model: /mnt/f/mergekit-my-test-gemma4/output001
    parameters:
      weight: 0.33
  - model: /mnt/e/text-generation-webui-1.14/user_data/models/k2-nano-karcher-100000
    parameters:
      weight: 0.33
parameters:
  normalize_weights: true
  overlap_alpha: 0.35
  ring_temperature: 4.0
  max_iter: 8
  tol: 1e-6
  eps: 1e-8
dtype: bfloat16

Downloads last month: 27

Safetensors

Model size

26B params

Tensor type

BF16

Model tree for win10/K2-nano-v1

Quantizations

2 models

Paper for win10/K2-nano-v1

Generalizing the Geometry of Model Merging Through Frechet Averages

Paper • 2604.27155 • Published May 7 • 1