Instructions to use CCCCCC/VPO-5B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CCCCCC/VPO-5B with Transformers:

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

pipe = pipeline("text-generation", model="CCCCCC/VPO-5B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("CCCCCC/VPO-5B")
model = AutoModelForCausalLM.from_pretrained("CCCCCC/VPO-5B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.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(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use CCCCCC/VPO-5B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "CCCCCC/VPO-5B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CCCCCC/VPO-5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/CCCCCC/VPO-5B

SGLang

How to use CCCCCC/VPO-5B 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 "CCCCCC/VPO-5B" \
    --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": "CCCCCC/VPO-5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "CCCCCC/VPO-5B" \
        --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": "CCCCCC/VPO-5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use CCCCCC/VPO-5B with Docker Model Runner:
```
docker model run hf.co/CCCCCC/VPO-5B
```

nielsr HF Staff commited on Sep 3, 2025

Commit

aee1d5b

verified ·

1 Parent(s): 9a43e1b

Improve model card: Remove non-standard metadata, shorten title, add citation and tags

Browse files

This PR improves the model card by:

- Removing the non-standard `inference: false` tag from the metadata.
- Shortening the main heading in the content for better readability.
- Adding the BibTeX citation for proper attribution.
- Adding `tags` metadata to better describe the model's function (prompt optimization for text-to-video generation).

The existing links to the paper, code repository, and sample usage are correctly preserved.

Files changed (1) hide show

README.md +13 -9

README.md CHANGED Viewed

@@ -1,19 +1,20 @@
 ---
 language:
 - en
-license: apache-2.0
-inference: false
 library_name: transformers
 pipeline_tag: text-generation
 ---
-<h1>VPO: Aligning Text-to-Video Generation Models with Prompt Optimization</h1>
 - **Repository:** https://github.com/thu-coai/VPO
 - **Paper:** [VPO: Aligning Text-to-Video Generation Models with Prompt Optimization](https://huggingface.co/papers/2503.20491)
 - **Data:** https://huggingface.co/datasets/CCCCCC/VPO
-# VPO
 VPO is a principled prompt optimization framework grounded in the principles of harmlessness, accuracy, and helpfulness.
 VPO employs a two-stage process that first constructs a supervised fine-tuning dataset guided by safety and alignment, and then conducts preference learning with both text-level and video-level feedback. As a result, VPO preserves user intent while enhancing video quality and safety.
@@ -82,9 +83,12 @@ print(resp)
 ```
 See our [Github Repo](https://github.com/thu-coai/VPO) for more detailed usage (e.g. Inference with Vllm).
-<!-- ## Citation
-If you find our model is useful in your work, please cite it with:
 ```
-``` -->

 ---
 language:
 - en
 library_name: transformers
+license: apache-2.0
 pipeline_tag: text-generation
+tags:
+- text-to-video
+- prompt-optimization
 ---
+<h1>VPO</h1>
 - **Repository:** https://github.com/thu-coai/VPO
 - **Paper:** [VPO: Aligning Text-to-Video Generation Models with Prompt Optimization](https://huggingface.co/papers/2503.20491)
 - **Data:** https://huggingface.co/datasets/CCCCCC/VPO
 VPO is a principled prompt optimization framework grounded in the principles of harmlessness, accuracy, and helpfulness.
 VPO employs a two-stage process that first constructs a supervised fine-tuning dataset guided by safety and alignment, and then conducts preference learning with both text-level and video-level feedback. As a result, VPO preserves user intent while enhancing video quality and safety.
 ```
 See our [Github Repo](https://github.com/thu-coai/VPO) for more detailed usage (e.g. Inference with Vllm).
+## Citation
 ```
+@article{cheng2025vpo,
+  title={Vpo: Aligning text-to-video generation models with prompt optimization},
+  author={Cheng, Jiale and Lyu, Ruiliang and Gu, Xiaotao and Liu, Xiao and Xu, Jiazheng and Lu, Yida and Teng, Jiayan and Yang, Zhuoyi and Dong, Yuxiao and Tang, Jie and others},
+  journal={arXiv preprint arXiv:2503.20491},
+  year={2025}
+}
+```