Hi @deneland
probably due to my ignorance of label studio, but on following these steps:

python3 -m venv venv
source venv/bin/activate
pip install streamlit-labelstudio
venv/bin/streamlit run app.py

The app raises exception:

StreamlitAPIException: No such component directory: '/Users/robin/Github/streamlit-labelstudio/streamlit_labelstudio/frontend/build'

File "/Users/robin/Github/streamlit-labelstudio/app.py", line 2, in <module>
    from streamlit_labelstudio import st_labelstudio
File "/Users/robin/Github/streamlit-labelstudio/streamlit_labelstudio/__init__.py", line 38, in <module>
    _component_func = components.declare_component("streamlit_labelstudio", path=build_dir)

I guess this might be due to Label Studio not being running?

It sometimes crops out the submit button, it loads only half the picture sometimes, and recently just I run the streamlit file and the label studio front end didn't even load (the screen was blank)

If i wish the text to be loaded with annotation when the component is displayed, i supposed the completions can help to pre-load with annotations right? Can I ask what if the format of accepted variable? I am not very familiar with front-end tech, couldnt figure out from the source code.

task = {
  "completions": [
    
    ],
  
  "predictions": [
        ],
  "data": {
    "text": "To have faith is to trust yourself to the water",
    'id': 20
    
},
}

Hi,
First of all, thanks for this great job!
To take full advantage of its potential, can you please update the ReadMe to guide the user on how to deploy your component in share.streamlit.io?
Thank you very much

hey,
I was running the app.py code. I'm totally new to streamlit and labelstudio, was wondering if there are any sample apps using this streamlit-labelstudio component that let the user upload multiple images.

hey @deneland,

this component looks very promising to me as i work with label stuido before.
However, i do not really get the benefit yet of wrapping it onto streamlit.

Cloud you provide an deployed streamlit app showing how it works?

Best regards
Chris

Hi,
I found returned bounding boxes' coordinates are not adapted to original image, and I don't know how to get the resize ratio between the displayed image and the original image. Could you please tell me the way to get the exact coordinates of the bounding boxes? Thanks a lot.

Hello, I am wondering if accessing to state variables during the labeling job is possible. For example, I want to access the label data - x, y, width, label,... - before pressing the Submit button.

How to add multiple annotation data