Instructions to use microsoft/OmniParser with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use microsoft/OmniParser with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="microsoft/OmniParser")# Load model directly from transformers import AutoProcessor, AutoModelForVisualQuestionAnswering processor = AutoProcessor.from_pretrained("microsoft/OmniParser") model = AutoModelForVisualQuestionAnswering.from_pretrained("microsoft/OmniParser") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use microsoft/OmniParser with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "microsoft/OmniParser" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "microsoft/OmniParser", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/microsoft/OmniParser
- SGLang
How to use microsoft/OmniParser 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 "microsoft/OmniParser" \ --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": "microsoft/OmniParser", "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 "microsoft/OmniParser" \ --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": "microsoft/OmniParser", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use microsoft/OmniParser with Docker Model Runner:
docker model run hf.co/microsoft/OmniParser
No examples in readme
Just an example or two would be ideal as I'm not sure if the model is rendering json and in what kind of schema.
@iiBLACKii - thank you for your notebook!
could you please elaborate on the blip2 model import?
Where does the icon_caption_blip2 model come from? Here on the huggingface repo are two bin files pytorch_model-00001-of-00002.bin and pytorch_model-00002-of-00002.bin.
In your notebook you are using this line:caption_model_processor = get_caption_model_processor(model_name="blip2", model_name_or_path="/content/drive/MyDrive/OmniParser/icon_caption_blip2", device="cuda")
On my end that loads the salesforce blip2 model & looks locally for the icon_caption_blip2 model - thats where i get the error:OSError: Incorrect path_or_model_id: '/Omniparser/icon_caption_blip2'. Please provide either the path to a local folder or the repo_id of a model on the Hub.
Hi @franperic , it seems the folder structure is incorrect. you can refer to the documentation: https://github.com/microsoft/OmniParser?tab=readme-ov-file#install, and see if it helps. Thanks.