Instructions to use yifanouyang/Lumen-Vision-460M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use yifanouyang/Lumen-Vision-460M with Transformers:

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

pipe = pipeline("image-text-to-text", model="yifanouyang/Lumen-Vision-460M", trust_remote_code=True)
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 AutoModel
model = AutoModel.from_pretrained("yifanouyang/Lumen-Vision-460M", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use yifanouyang/Lumen-Vision-460M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "yifanouyang/Lumen-Vision-460M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "yifanouyang/Lumen-Vision-460M",
		"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/yifanouyang/Lumen-Vision-460M

SGLang

How to use yifanouyang/Lumen-Vision-460M 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 "yifanouyang/Lumen-Vision-460M" \
    --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": "yifanouyang/Lumen-Vision-460M",
		"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 "yifanouyang/Lumen-Vision-460M" \
        --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": "yifanouyang/Lumen-Vision-460M",
		"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 yifanouyang/Lumen-Vision-460M with Docker Model Runner:
```
docker model run hf.co/yifanouyang/Lumen-Vision-460M
```

Lumen Vision 460M

A compact English single-image VLM for image questions and document experiments. 460M parameters, with a complete 2.03 GB FP32 checkpoint.

Code & quick start · Local / offline setup · Training

hf download yifanouyang/Lumen-Vision-460M --local-dir checkpoints/lumen

Local workflow

The GitHub CLI supports individual images or folders, JSONL answers and standalone HTML reports. It uses local implementation code, pinned checkpoints, orientation correction and image-token checks.

Connector adaptation freezes both backbones and saves a separate adapter. CPU inference and a one-step CPU connector smoke run were checked. Released base weights are unchanged; no new benchmark scores are reported.

English-first, single-image input. Hallucinations and counting errors remain possible. Direct Transformers Hub loading requires reviewing the repository code and enabling trust_remote_code=True; the GitHub CLI uses its local implementation.

License

Apache-2.0 · Source · Model reference · Notices

Downloads last month: 42

Safetensors

Model size

0.5B params

Tensor type

F32

Model tree for yifanouyang/Lumen-Vision-460M

Base model

lusxvr/nanoVLM-460M-8k

Finetuned

qvac/VisionPsy-Nano-460M

Finetuned

(3)

this model