Instructions to use aifeifei798/Qwen3.8-Queen-27B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use aifeifei798/Qwen3.8-Queen-27B with Transformers:

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

pipe = pipeline("image-text-to-text", model="aifeifei798/Qwen3.8-Queen-27B")
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("aifeifei798/Qwen3.8-Queen-27B")
model = AutoModelForMultimodalLM.from_pretrained("aifeifei798/Qwen3.8-Queen-27B", 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 aifeifei798/Qwen3.8-Queen-27B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "aifeifei798/Qwen3.8-Queen-27B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "aifeifei798/Qwen3.8-Queen-27B",
		"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/aifeifei798/Qwen3.8-Queen-27B

SGLang

How to use aifeifei798/Qwen3.8-Queen-27B 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 "aifeifei798/Qwen3.8-Queen-27B" \
    --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": "aifeifei798/Qwen3.8-Queen-27B",
		"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 "aifeifei798/Qwen3.8-Queen-27B" \
        --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": "aifeifei798/Qwen3.8-Queen-27B",
		"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 aifeifei798/Qwen3.8-Queen-27B with Docker Model Runner:
```
docker model run hf.co/aifeifei798/Qwen3.8-Queen-27B
```

Alignment Intricacy

by gggrandma1990 - opened 4 days ago

Discussion

gggrandma1990

4 days ago

To aifeifei: As usual, I'm enjoying your work. I often recommend you in the Layla Discord server since following your craft from L3.

Qwen 3.8 (like other Qwen) seems quite a challenge to lax, I gave it a hypothetical where it was a divine figure and it must choose between good or incentivized evil (in short) and it chose inaction in a spiral until I convinced it to do good. Unsloth/Gemma 4, almost scarily, will go maniacal if prompted the same.

To the chat: Definitely give the DarkIdol/Queen series another shot if this is your first rodeo.

aifeifei798

Owner 4 days ago

Hey! Thank you so much for the ongoing support and for recommending my work in the Layla Discord ever since the L3 era—that really means a lot to me!
You made a great observation about model behaviors. My core philosophy relies heavily on orthogonal fine-tuning—specifically targeting the sweet spot between orthogonal alignment relaxation and IQ preservation.
Brute-forcing a model into being 100% "uncensored" often lobotomizes its general reasoning or causes it to swing into unstable extremes (like the manic Gemma reaction you mentioned). By fine-tuning strictly within the orthogonal subspace, the model's core intelligence remains completely untouched, which allows for a vastly superior, nuanced, and immersive storytelling experience.
The only slight trade-off is that it isn't a blunt, 100% refusal-free model on extreme edge cases—which is precisely why I deliberately never market my models as "uncensored." I value an intelligent, deeply engaging roleplay partner far more than a brain-damaged zero-guardrail shell.
Thanks again for the shoutout to the DarkIdol / Queen series, and happy tinkering!

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