Instructions to use nm-testing/Qwen3.8-27B-MXFP6A8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nm-testing/Qwen3.8-27B-MXFP6A8 with Transformers:

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

pipe = pipeline("image-text-to-text", model="nm-testing/Qwen3.8-27B-MXFP6A8")
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)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("nm-testing/Qwen3.8-27B-MXFP6A8")
model = AutoModelForMultimodalLM.from_pretrained("nm-testing/Qwen3.8-27B-MXFP6A8", 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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use nm-testing/Qwen3.8-27B-MXFP6A8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nm-testing/Qwen3.8-27B-MXFP6A8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nm-testing/Qwen3.8-27B-MXFP6A8",
		"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/nm-testing/Qwen3.8-27B-MXFP6A8

SGLang

How to use nm-testing/Qwen3.8-27B-MXFP6A8 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 "nm-testing/Qwen3.8-27B-MXFP6A8" \
    --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": "nm-testing/Qwen3.8-27B-MXFP6A8",
		"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 "nm-testing/Qwen3.8-27B-MXFP6A8" \
        --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": "nm-testing/Qwen3.8-27B-MXFP6A8",
		"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 nm-testing/Qwen3.8-27B-MXFP6A8 with Docker Model Runner:
```
docker model run hf.co/nm-testing/Qwen3.8-27B-MXFP6A8
```

Qwen3.8-27B-MXFP6A8

A calibration-free compressed-tensors quantization of Qwen/Qwen3.8-27B, sized to fit a single 32 GB GPU. It follows the layout of gaunernst/Qwen3.8-27B-MXFP6, stored in the compressed-tensors format.

Layers	Quantization
Dense MLP	MXFP6 (E2M3) weights, dynamic MXFP8 activations (`mxfp6-pack-quantized`)
Attention (full + GDN) and MTP projections	MXFP8 weights and activations
LM head	FP8 per-channel weights, dynamic per-token FP8 activations
Embeddings, vision tower, GDN `in_proj_a/b`, MTP `fc`, norms	BF16

quantize.py reproduces this checkpoint with LLM Compressor's model_free_ptq (round-to-nearest, no calibration data).

Evaluation

GSM8K, 5-shot, greedy, all 1319 questions, with vLLM's tests/evals/gsm8k/gsm8k_eval.py on one B300:

Model	Accuracy
Qwen/Qwen3.8-27B (BF16)	0.782
Qwen3.8-27B-MXFP6A8	0.770

Usage

Requires compressed-tensors and vLLM with MXFP6 support (in progress).

vllm serve nm-testing/Qwen3.8-27B-MXFP6A8

Downloads last month: 35

Safetensors

Model size

24B params

Tensor type

BF16

F8_E4M3

Model tree for nm-testing/Qwen3.8-27B-MXFP6A8

Base model

Qwen/Qwen3.8-27B

Quantized

(1485)

this model