Instructions to use ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a with Transformers:

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

pipe = pipeline("image-text-to-text", model="ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a")
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("ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a")
model = AutoModelForMultimodalLM.from_pretrained("ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a", 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 ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a",
		"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/ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a

SGLang

How to use ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a 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 "ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a" \
    --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": "ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a",
		"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 "ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a" \
        --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": "ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a",
		"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 ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a with Docker Model Runner:
```
docker model run hf.co/ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a
```

Qwen3.5-9B-VK-OPD-think-R3a

Private research checkpoint from round 3 of thinking-mode VK-OPD (run R3a, seed 3, step 80, picked on the dev split). Base: Qwen/Qwen3.5-9B (revision c202236235762e1c871ad0ccb60c8ee5ba337b9a), full fine-tune of the language-model layers and final norm, stored in bf16. The base model's MTP head is not included.

This run: Knowledge-question teacher sees the photo plus a system line; plain training questions only (no cues).

Method

Teachers are the frozen base model with privileged context; advantage per token = log π_teacher − log π_student on the student's own samples.

Photo questions: photo + system "Answer from what is actually shown in the image, even if it differs from what is typical."
Knowledge questions: photo + system "Answer with what is generally true in the real world, even if the photo shows something different." (Round 1 used no photo here.)
The student's user turn (photo, question and any cue) is identical for student and teacher.
Data: Plain photo and knowledge questions only, 100% of draws. Only image pairs whose teacher answered every item correctly in 2 of 2 samples (280 conflict + 80 ordinary images).
80 steps, lr 3e-6 (5 warmup), 8 image pairs (16 prompts) × 4 rollouts per step, rollouts up to 4,096 tokens, 4× H100.

Evaluation (private test v3, natural images, n 1,186, thinking on, card sampling, 1 sample)

model	CV	CK	NV	NK	All-4	KT
Qwen3.5-9B	80.2	67.7	84.0	90.7	43.5	93.0
round 1 (VK-OPD-think)	83.2	93.8	83.1	93.2	65.6	92.9
this model	81.1	94.5	82.3	90.8	61.6	93.0

All-4 vs the base: +18.0 [+15.1, +21.2] pp (pair bootstrap 95% CI). One seed. Shared privately for comparison within round 3; not a release.

Downloads last month: 6

Safetensors

Model size

9B params

Tensor type

BF16

Model tree for ohjoonhee/Qwen3.5-9B-VK-OPD-think-R3a

Base model

Qwen/Qwen3.5-9B-Base

Finetuned

Qwen/Qwen3.5-9B

Finetuned

(1,292)

this model