Instructions to use zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30 with Transformers:

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

pipe = pipeline("image-text-to-text", model="zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30")
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("zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30")
model = AutoModelForMultimodalLM.from_pretrained("zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30", 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 zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30",
		"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/zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30

SGLang

How to use zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30 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 "zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30" \
    --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": "zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30",
		"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 "zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30" \
        --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": "zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30",
		"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 zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30 with Docker Model Runner:
```
docker model run hf.co/zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30
```

Vision-OPD Qwen2.5-VL-7B RandomDrop5 Step30

Qwen2.5-VL-7B-Instruct trained for 30 Vision-OPD optimizer steps. The student retains a replayable random 5% subset of post-merge visual tokens, while the EMA teacher uses the full visual-token sequence.

Training uses Vision-OPD-6K with 96 prompts and 8 on-policy rollouts per step, a learning rate of 2e-6, and an EMA teacher update rate of 0.05.

Standard Transformers/vLLM loads the full-token model. RandomDrop5 inference requires the pruning-aware serving code in prune-opd.

Training data: Vision-OPD-6K.

Downloads last month: 11

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for zhuqiang/Vision-OPD-Qwen2.5-VL-7B-Instruct-RandomDrop5-Step30

Base model

Qwen/Qwen2.5-VL-7B-Instruct

Finetuned

(1186)

this model