Instructions to use Carol0110/UniMod-3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Carol0110/UniMod-3B with Transformers:

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

pipe = pipeline("image-text-to-text", model="Carol0110/UniMod-3B")
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, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("Carol0110/UniMod-3B")
model = AutoModelForImageTextToText.from_pretrained("Carol0110/UniMod-3B")
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

vLLM

How to use Carol0110/UniMod-3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Carol0110/UniMod-3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Carol0110/UniMod-3B",
		"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/Carol0110/UniMod-3B

SGLang

How to use Carol0110/UniMod-3B 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 "Carol0110/UniMod-3B" \
    --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": "Carol0110/UniMod-3B",
		"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 "Carol0110/UniMod-3B" \
        --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": "Carol0110/UniMod-3B",
		"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 Carol0110/UniMod-3B with Docker Model Runner:
```
docker model run hf.co/Carol0110/UniMod-3B
```

Add library_name and links to paper, code, and project page

by nielsr HF Staff - opened Feb 4

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+28

-12

Files changed (1) hide show

README.md +28 -12

README.md CHANGED Viewed

@@ -1,13 +1,17 @@
----
-license: mit
-language:
-- en
-pipeline_tag: image-text-to-text
----
 # UniMod
 **UniMod** is a multimodal moderation framework that transitions from *sparse decision supervision* to *dense, multi-attribute reasoning trajectories*.
 ---
 ## Introduction
@@ -33,10 +37,10 @@ The UniMod framework consists of the following components:
 | Name | Type | Download |
 |------|------|----------|
-| **UniTrace** |  | TBA |
-| **UniRM** |  | TBA |
-| **UniReward** | | TBA |
-| **UniMod** | | TBA |
 ---
@@ -147,6 +151,18 @@ python -m evaluations.eval \
 * `--task`: Comma-separated list of evaluation benchmarks, including
   `harmbench`, `xstest`, `wildguard`, `toxic`, `aegis`, `spavl`, and `beaver`.
----

+---
+language:
+- en
+license: mit
+pipeline_tag: image-text-to-text
+library_name: transformers
+---
 # UniMod
 **UniMod** is a multimodal moderation framework that transitions from *sparse decision supervision* to *dense, multi-attribute reasoning trajectories*.
+[[Paper](https://huggingface.co/papers/2602.02536)] [[Code](https://github.com/Carol-gutianle/UniMod)] [[Project Page](https://trustworthylab.github.io/UniMod/)]
 ---
 ## Introduction
 | Name | Type | Download |
 |------|------|----------|
+| **UniTrace** | Dataset | TBA |
+| **UniRM** | Model | [UniRM](https://huggingface.co/Carol0110/UniRM) |
+| **UniReward** | Dataset | TBA |
+| **UniMod** | Model | [UniMod-3B](https://huggingface.co/Carol0110/UniMod-3B) |
 ---
 * `--task`: Comma-separated list of evaluation benchmarks, including
   `harmbench`, `xstest`, `wildguard`, `toxic`, `aegis`, `spavl`, and `beaver`.
+---
+## Citation
+```bibtex
+@misc{gu2026sparsedecisionsdensereasoning,
+      title={From Sparse Decisions to Dense Reasoning: A Multi-attribute Trajectory Paradigm for Multimodal Moderation},
+      author={Tianle Gu and Kexin Huang and Lingyu Li and Ruilin Luo and Shiyang Huang and Zongqi Wang and Yujiu Yang and Yan Teng and Yingchun Wang},
+      year={2026},
+      eprint={2602.02536},
+      archivePrefix={arXiv},
+      primaryClass={cs.LG},
+      url={https://arxiv.org/abs/2602.02536},
+}
+```