Instructions to use hustvl/InfiniteVL-LongSFT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use hustvl/InfiniteVL-LongSFT with Transformers:

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

pipe = pipeline("image-text-to-text", model="hustvl/InfiniteVL-LongSFT", 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)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("hustvl/InfiniteVL-LongSFT", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use hustvl/InfiniteVL-LongSFT with vLLM:

Install from pip and serve model

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

SGLang

How to use hustvl/InfiniteVL-LongSFT 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 "hustvl/InfiniteVL-LongSFT" \
    --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": "hustvl/InfiniteVL-LongSFT",
		"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 "hustvl/InfiniteVL-LongSFT" \
        --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": "hustvl/InfiniteVL-LongSFT",
		"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 hustvl/InfiniteVL-LongSFT with Docker Model Runner:
```
docker model run hf.co/hustvl/InfiniteVL-LongSFT
```

Improve model card: Update arXiv link and add comprehensive details from GitHub

by nielsr HF Staff - opened Dec 12, 2025

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+154

-7

Improve model card: Update arXiv link and add comprehensive details from GitHub874cdd1a

nielsr

Dec 12, 2025

Hello team,

I've opened this PR to enhance the model card for the InfiniteVL model. The updates aim to provide more comprehensive and accurate information for users.

Key changes include:

Corrected arXiv paper link: The previous placeholder https://arxiv.org/abs/2502.xxxxx has been updated to the correct link https://arxiv.org/abs/2512.08829, consistent with the paper info and GitHub repository.
Added Hugging Face badge: A Hugging Face badge has been included in the header to improve navigation and visibility of the model on the Hub.
Enriched content from GitHub README: Several crucial sections from the project's GitHub README have been integrated into the model card, including:
- News
- Table of Contents (updated to reflect the model card structure)
- Architecture (including relevant images with absolute GitHub URLs)
- Training Strategy (including relevant images with absolute GitHub URLs)
- Performance (including relevant images with absolute GitHub URLs)
- Qualitative Analysis & Visualization (including relevant images with absolute GitHub URLs)
- Contact
Detailed Advanced Usage: The brief "Advanced Usage (Cuda Graph)" section has been replaced with the more detailed explanation and code snippets from the GitHub README.
Refined Metadata Tags: Removed the redundant image-text-to-text tag from the tags list, as it is already present in pipeline_tag.

These changes ensure the model card is more informative, accurate, and easier to navigate for anyone exploring InfiniteVL.
All relative image paths from the GitHub README have been converted to absolute https://github.com/hustvl/InfiniteVL/raw/main/assets/... URLs.

HongyuanTao changed pull request status to merged Dec 12, 2025

HongyuanTao

HUST Vision Lab org Dec 12, 2025

Thank you for the improvements! I really appreciate the help with the model card.

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