Instructions to use Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw with Transformers:

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

pipe = pipeline("image-text-to-text", model="Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw")
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("Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw")
model = AutoModelForMultimodalLM.from_pretrained("Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw", 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 Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw",
		"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/Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw

SGLang

How to use Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw 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 "Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw" \
    --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": "Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw",
		"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 "Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw" \
        --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": "Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw",
		"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 Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw with Docker Model Runner:
```
docker model run hf.co/Billyjack3791/Qwen3.6-35B-A3B-EXL3-5.0bpw
```

Qwen3.6-35B-A3B-EXL3-5.0bpw

EXL3 quantized version of Qwen/Qwen3.6-35B-A3B at 5.0 bits per weight. Mixture-of-Experts architecture.

Quantization Details

Parameter	Value
Source model	Qwen/Qwen3.6-35B-A3B
Quantization method	EXL3
EXL3 version	v0.0.30
Bits per weight	5.0
Codebook	mcg
Calibration	250 rows x 2048 columns
Architecture	MoE (256 experts, 8 active)

Hardware Used

Component	Specification
CPU	Intel Core i7-9800X @ 3.80GHz (8C/16T)
Motherboard	ASUS WS X299 SAGE
RAM	32GB DDR4-2666
GPU(s)	6x NVIDIA GeForce RTX 3070 8GB
Storage	MSI M480 PRO 1TB NVMe
OS	Ubuntu 24.04.4 LTS
Driver	NVIDIA 580.173.02

Orchestration: Quantization jobs were orchestrated from a dedicated control node (Intel i9-7920X, 62GB RAM, Quadro P5000 + RTX 3070 Ti + 2x RTX 3080) over a dedicated 10Gb interconnect to the compute node above.

File Sizes

File	Size
model-00001-of-00003.safetensors	8.45 GB
model-00002-of-00003.safetensors	8.50 GB
model-00003-of-00003.safetensors	6.71 GB
Total	23.70 GB

Notes

Custom quantization of the Qwen3.6 MoE variant - not available from other sources at this precision
Uses mcg codebook (optimal for MoE architectures)
Clean quantization with no hardware errors

Downloads last month: 5

Safetensors

Model size

12B params

Tensor type

BF16

F16

I16