Instructions to use selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB with Transformers:

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

pipe = pipeline("image-text-to-text", model="selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB", trust_remote_code=True)
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("selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB", trust_remote_code=True)
model = AutoModelForMultimodalLM.from_pretrained("selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB", trust_remote_code=True)
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 selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB",
		"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/selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB

SGLang

How to use selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB 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 "selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB" \
    --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": "selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB",
		"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 "selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB" \
        --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": "selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB",
		"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 selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB with Docker Model Runner:
```
docker model run hf.co/selode-ai/Qwen-3.6-35B-A3B-VRAP-4-bit-AWQ-21.2GB
```

Thank you for this model

by Abalagaev - opened 23 days ago

Discussion

Abalagaev

23 days ago

•

edited 23 days ago

Althought it is broken, Claude managed somehow to fix it, and now I do have 16k context on single RTX 3090 that allows me to reach 400ms latency on scene-understanding and robotic action via vLLM, that allows me to use System-2-only skills on LeKiwi. What a time to be alive!

selode-ai

Owner 3 days ago

•

edited 3 days ago

Since we started working on this methodology, there have been several updates to tensor shaping, that has made the VLLM active version no longer working.

Thank you very much for persisting with claude code and we're very happy that you saw value in this release, as we have more in the pipeline 😀

As a development team, we maintain a fork of SGLang current with latest releases. Consequently, we've just uploaded as of a few minutes ago full instructions for running the latest release on the latest version of SGLang, requiring just one file to be copied acrossl, it should be more performant than running on VLLM, as a baseline, based on our initial testing and subsequent enhancements.

Enjoy, and we look forward to promoting more releases in the future.

Selode Team

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