Instructions to use YipengZhang/LLaVA-UHD-v2-Vicuna-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use YipengZhang/LLaVA-UHD-v2-Vicuna-7B with Transformers:

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

pipe = pipeline("image-text-to-text", model="YipengZhang/LLaVA-UHD-v2-Vicuna-7B")

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("YipengZhang/LLaVA-UHD-v2-Vicuna-7B", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use YipengZhang/LLaVA-UHD-v2-Vicuna-7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "YipengZhang/LLaVA-UHD-v2-Vicuna-7B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "YipengZhang/LLaVA-UHD-v2-Vicuna-7B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/YipengZhang/LLaVA-UHD-v2-Vicuna-7B

SGLang

How to use YipengZhang/LLaVA-UHD-v2-Vicuna-7B 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 "YipengZhang/LLaVA-UHD-v2-Vicuna-7B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "YipengZhang/LLaVA-UHD-v2-Vicuna-7B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "YipengZhang/LLaVA-UHD-v2-Vicuna-7B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "YipengZhang/LLaVA-UHD-v2-Vicuna-7B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use YipengZhang/LLaVA-UHD-v2-Vicuna-7B with Docker Model Runner:
```
docker model run hf.co/YipengZhang/LLaVA-UHD-v2-Vicuna-7B
```

LLaVA-UHD v2 Model Card

Model details

Model type: LLaVA-UHD v2, an advanced MLLM centered around a Hierarchical window transformer that enables capturing diverse visual granularity by constructing and integrating a high resolution feature pyramid.

Model date: LLaVA-UHD v2 was trained in November 2024.

Base LLM Model: lmsys/vicuna-7b-v1.5

Paper or resources for more information: https://github.com/thunlp/LLaVA-UHD

License

Where to send questions or comments about the model: https://github.com/thunlp/LLaVA-UHD/issues

Intended use

Primary intended uses: The primary use of LLaVA-UHD v2 is research on large multimodal models and chatbots.

Primary intended users: The primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence.

Training dataset

VDIM Pretrain: MS-COCO stuff 2017
Pretrain: LLaVA-Pretrain 558K (filtered image-text pairs from LAION/CC/SBU, captioned by BLIP.)
SFT: 858k-mixed dataset in https://huggingface.co/datasets/YipengZhang/LLaVA-UHD-v2-SFT-Data

Citation

If you find LLaVA-UHD v2 useful for your research and applications, please cite using this BibTeX:

@article{zhang2024llavauhdv2,
  title={LLaVA-UHD v2: an MLLM Integrating High-Resolution Feature Pyramid via Hierarchical Window Transformer},
  author={Yipeng Zhang and Yifan Liu and Zonghao Guo and Yidan Zhang and Xuesong Yang and Chi Chen and Jun Song and Bo Zheng and Yuan Yao and Zhiyuan Liu and Tat-Seng Chua and Maosong Sun},
  journal={arXiv preprint arXiv:2412.13871},
  year={2024}
}

Downloads last month: 9

Safetensors

Model size

8B params

Tensor type

BF16

Inference Providers NEW

Image-Text-to-Text

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for YipengZhang/LLaVA-UHD-v2-Vicuna-7B

Base model

lmsys/vicuna-7b-v1.5

Finetuned

(65)

this model

Dataset used to train YipengZhang/LLaVA-UHD-v2-Vicuna-7B

Paper for YipengZhang/LLaVA-UHD-v2-Vicuna-7B

LLaVA-UHD v2: an MLLM Integrating High-Resolution Feature Pyramid via Hierarchical Window Transformer

Paper • 2412.13871 • Published Dec 18, 2024 • 18